290 7 5MB
English Pages 960 [952] Year 2004
Table of Contents
Video Demystified
Video Demystified
i
ii
Table of Contents
This page intentionally left blank
Table of Contents
Video Demystified A Handbook for the Digital Engineer Fourth Edition
by Keith Jack
AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Newnes is an imprint of Elsevier
iii
iv
Table of Contents
Newnes is an imprint of Elsevier 200 Wheeler Road, Burlington, MA 01803, USA Linacre House, Jordan Hill, Oxford OX2 8DP, UK Copyright © 2005, Elsevier Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Permissions may be sought directly from Elsevier's Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting “Customer Support” and then “Obtaining Permissions.” Recognizing the importance of preserving what has been written, Elsevier prints its books on acid-free paper whenever possible. Librar y of Congress Cataloging-in-Publication Data (Application submitted.) British Librar y Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: 0-7506-7822-4 For information on all Newnes publications visit our Web site at www.books.elsevier.com 04 05 06 07 08 09 10 9 8 7 6 5 4 3 2 1 Printed in the United States of America
Contents
Contents
v
Contents About the Author
xxi
What's on the CD-ROM? Chapter 1 •
Introduction
xxiii 1
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Organization Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Chapter 2 •
Introduction to Video
6
Analog vs. Digital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Video Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Digital Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Video Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Video Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Standard Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Enhanced Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 High Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Audio and Video Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Application Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 DVD Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Digital Media Adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Digital Television Set-top Boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
v
vi
Contents
Chapter 3 •
Color Spaces
15
RGB Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 YUV Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 YIQ Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 YCbCr Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 RGB - YCbCr Equations: SDTV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 RGB - YCbCr Equations: HDTV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4:4:4 YCbCr Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4:2:2 YCbCr Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4:1:1 YCbCr Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 4:2:0 YCbCr Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 PhotoYCC Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 HSI, HLS and HSV Color Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Chromaticity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Non-RGB Color Space Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Gamma Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Constant Luminance Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Chapter 4 •
Video Signals Overview
35
Digital Component Video Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Coding Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 480i and 480p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Interlaced Analog Composite Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Interlaced Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Progressive Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Interlaced Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Progressive Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 576i and 576p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Interlaced Analog Composite Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Interlaced Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Progressive Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Interlaced Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Progressive Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Contents
720p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progressive Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progressive Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080i and 1080p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interlaced Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progressive Analog Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interlaced Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progressive Digital Component Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Video Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 5 •
Analog Video Interfaces
vii
54 54 54 57 57 57 60 62 62 65
66
S-Video Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCART Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SDTV RGB Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 IRE Blanking Pedestal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 IRE Blanking Pedestal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HDTV RGB Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copy Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SDTV YPbPr Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VBI Data for 480p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VBI Data for 576p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HDTV YPbPr Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VBI Data for 720p Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VBI Data for 1080i Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copy Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-Connector Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Pro-Video Analog Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VGA Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
66 67 69 69 72 73 75 75 80 83 86 90 90 91 91 92 92 92
viii
Contents
Chapter 6 •
Digital Video Interfaces
100
Pro-Video Component Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Video Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Ancillary Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 25-pin Parallel Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 93-pin Parallel Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Serial Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 SDTV—Interlaced . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 SDTV—Progressive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 HDTV—Interlaced . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 HDTV—Progressive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Pro-Video Composite Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 NTSC Video Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 PAL Video Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Ancillary Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 25-pin Parallel Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Serial Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Pro-Video Transport Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Serial Data Transport Interface (SDTI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 High Data-Rate Serial Data Transport Interface (HD-SDTI) . . . . . . . . . . . . . . . . . . . . . . . 156 IC Component Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 “BT.601” Video Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Video Module Interface (VMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 “BT.656” Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Zoomed Video Port (ZV Port) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Video Interface Port (VIP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Consumer Component Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Digital Visual Interface (DVI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 High Definition Multimedia Interface (HDMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Digital Flat Panel (DFP) Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Open LVDS Display Interface (OpenLDI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Gigabit Video Interface (GVIF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Consumer Transport Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 USB 2.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Ethernet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 IEEE 1394 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Contents
Chapter 7 •
Digital Video Processing
ix
202
Rounding Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SDTV - HDTV YCbCr Transforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4:4:4 to 4:2:2 YCbCr Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Enhancement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brightness, Contrast, Saturation (Color) and Hue (Tint) . . . . . . . . . . . . . . . . . . . . . . . . . Color Transient Improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sharpness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Mixing and Graphics Overlay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luma and Chroma Keying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pixel Dropping and Duplication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linear Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anti-Aliased Resampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Scaling Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scan Rate Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frame or Field Dropping and Duplicating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temporal Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2:2 Pulldown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3:2 Pulldown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3:3 Pulldown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24:1 Pulldown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Noninterlaced-to-Interlaced Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scan Line Decimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vertical Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interlaced-to-Noninterlaced Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Mode: Intrafield Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Mode: Interfield Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Film Mode (using Inverse Telecine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frequency Response Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DCT-Based Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fixed Pixel Display Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Expanded Color Reproduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detail Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-uniform Quantization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scaling and Deinterlacing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
203 204 205 208 208 210 210 212 219 231 232 232 232 235 235 240 242 248 248 249 249 249 249 251 251 251 253 255 255 256 260 260 260 260 261 261 264
x
Contents
Chapter 8 •
NTSC, PAL, and SECAM Overview
265
NTSC Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 Luminance Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 Color Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 Color Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 Composite Video Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Color Subcarrier Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 NTSC Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 RF Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 PAL Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 Luminance Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 Color Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Color Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Composite Video Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 PAL Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 RF Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 PALplus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 SECAM Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Luminance Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Color Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Color Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Composite Video Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 SECAM Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Video Test Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 VBI Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Timecode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 EIA–608 Closed Captioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354 Widescreen Signalling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 Teletext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 “Raw” VBI Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 “Sliced” VBI Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 NTSC/PAL Decoder Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Ghost Cancellation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Enhanced Television Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
Contents
Chapter 9 •
NTSC and PAL Digital Encoding and Decoding
xi
394
NTSC and PAL Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2× Oversampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Color Space Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luminance (Y) Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Color Difference Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analog Composite Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Color Subcarrier Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Horizontal and Vertical Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clean Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bandwidth-Limited Edge Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Level Limiting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Encoder Video Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genlocking Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alpha Channel Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NTSC and PAL Digital Decoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digitizing the Analog Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Y/C Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Color Difference Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luminance (Y) Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Color Space Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genlocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Timing Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Auto-Detection of Video Signal Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Y/C Separation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alpha Channel Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Decoder Video Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
395 395 395 399 402 410 413 417 421 422 423 423 427 428 428 428 431 431 435 438 440 442 450 452 452 464 467 471
xii
Contents
Chapter 10 •
H.261 and H.263
472
H.261 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472 Video Coding Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472 Video Bitstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Still Image Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 H.263 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Video Coding Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 Video Bitstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 Optional H.263 Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
Chapter 11 •
Consumer DV
519
Audio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521 IEC 61834 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521 SMPTE 314M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Audio Auxiliary Data (AAUX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 DCT Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 Macroblocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 Super Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 Video Auxiliary Data (VAUX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Digital Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536 IEEE 1394 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 SDTI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 100 Mbps DV Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Contents
Chapter 12 •
MPEG-1
xiii
543
MPEG vs. JPEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Audio Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sound Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Coding Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interlaced Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Encode Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coded Frame Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motion Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Bitstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sequence Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group of Pictures (GOP) Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Picture Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slice Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macroblock (MB) Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . System Bitstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISO/IEC 11172 Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pack Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . System Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Decoding 579 Fast Playback Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pause Mode Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reverse Playback Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Decode Postprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Real-World Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
543 544 545 545 546 546 547 547 547 549 550 552 553 554 555 555 555 559 560 561 562 566 574 574 574 575 577 579 579 579 579 580 580
xiv
Contents
Chapter 13 •
MPEG-2
581
Audio Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582 Video Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582 Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582 Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582 Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588 Video Coding Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 YCbCr Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 Coded Picture Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 Motion Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590 Macroblocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591 I Pictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591 P Pictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594 B Pictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 Video Bitstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 Video Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 Sequence Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 User Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600 Sequence Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600 Sequence Display Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602 Sequence Scalable Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Group of Pictures (GOP) Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607 Picture Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 Content Description Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609 Picture Coding Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615 Quant Matrix Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618 Picture Display Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620 Picture Temporal Scalable Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621 Picture Spatial Scalable Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622 Copyright Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623 Camera Parameters Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624 ITU-T ext. D Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624 Active Format Description (AFD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624 Slice Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624 Macroblock Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626 Block Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635 Motion Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636 PES Packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Contents
Program Stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pack Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . System Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program Stream Map (PSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program Stream Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transport Stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adaptation Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program Specific Information (PSI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program Association Table (PAT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program Map Table (PMT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transport Stream Description Table (TSDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conditional Access Table (CAT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Network Information Table (NIT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IPMP Control Information Table (ICIT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPEG-2 PMT/PSM Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPEG-4 PMT/PSM Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ARIB PMT Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ATSC PMT Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DVB PMT Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OpenCable PMT Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Closed Captioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EIA–708 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ATSC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digital Cable - SCTE 21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digital Cable - SCTE 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DVB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teletext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DVB VBI Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DVB EBU Teletext Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Widescreen Signaling (WSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subtitles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digital Cable Subtitles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DVB Subtitles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Enhanced Television Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intellectual Property Management and Protection (IPMP) . . . . . . . . . . . . . . . . . . . . . . . . . . . MPEG-4 over MPEG-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H.264 over MPEG-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SMPTE VC-9 over MPEG-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv
662 663 663 665 667 667 667 669 672 674 676 677 678 679 679 680 691 695 698 701 706 707 707 709 711 711 713 713 713 715 716 716 716 720 721 723 724 724 724
xvi
Contents
Data Broadcasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 Carousels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 IP Multicasting over MPEG-2 Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 Data Broadcasting Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728 Decoder Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729 Audio and Video Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729 Testing Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
Chapter 14 •
MPEG-4 and H.264
736
Audio Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 General Audio Object Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 Speech Object Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738 Synthesized Speech Object Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738 Synthesized Audio Object Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738 Visual Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739 YCbCr Color Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739 Visual Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739 MPEG-4 Part 2 Natural Visual Object Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739 MPEG-4 Part 2 Natural Visual Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741 Graphics Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741 Visual Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Visual Object Sequence (VS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Video Object (VO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Video Object Layer (VOL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Group of Video Object Plane (GOV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Video Object Plane (VOP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747 Object Description Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747 Object Descriptor (OD) Stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747 Object Content Information (OCI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Intellectual Property Management and Protection (IPMP) . . . . . . . . . . . . . . . . . . . . . . . . 749 Scene Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 BIFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Synchronization of Elementary Streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751 Sync Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751 DMIF Application Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752 Multiplexing of Elementary Streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752 FlexMux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Contents
MPEG-4 Over MPEG-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MP4 File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intellectual Property Management and Protection (IPMP) . . . . . . . . . . . . . . . . . . . . . . . . . . . MPEG-4 Part 10 (H.264) Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profiles and Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Coding Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Network Abstraction Layer (NAL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 15 •
ATSC Digital Television
752 753 753 753 754 754 759 759
760
Video Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Audio Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program and System Information Protocol (PSIP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Required Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optional Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Broadcasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Service Announcements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service Description Framework (SDF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Terrestrial Transmission Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-VSB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Audio Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modulation System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program and System Information Protocol (PSIP-E) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 16 •
xvii
OpenCable™ Digital Television
762 762 764 764 764 766 768 768 769 770 772 772 773 773 773 773 777
778
Video Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Audio Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In-Band System Information (SI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Required Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optional Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
780 780 780 781 782 784
xviii
Contents
Out-of-Band System Information (SI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788 In-Band Data Broadcasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Data Service Announcements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Service Description Framework (SDF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791 Conditional Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 Related Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 DOCSIS® . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 PacketCable™ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 Application Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Chapter 17 •
DVB Digital Television
796
Video Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 Audio Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 System Information (SI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 Required Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 Optional Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799 Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804 Teletext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Subtitles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Widescreen Signaling (WSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Data Broadcasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Transmission Format: Terrestrial (DVB-T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 Transmission Format: Cable (DVB-C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 Transmission Format: Satellite (DVB-S) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 Transmission Format: Satellite (DVB-S2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 Transmission Format: Mobile (DVB-H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 Conditional Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 Simulcrypt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813 Multicrypt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813 DVB Common Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813 Application Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Contents
Chapter 18 •
ISDB Digital Television
816
ISDB-S (Satellite) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISDB-C (Cable) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISBD-T (Terrestrial) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Video Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Audio Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Still Picture Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Graphics Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . System Information (SI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Captioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Broadcasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 19 •
IPTV
Index
897
817 817 818 818 818 819 819 819 819 821 829 829 830 830
831
Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multicasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RTSP-Based Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RTSP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RTCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RSVP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Broadcast over IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conditional Access (DRM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 20 •
xix
Glossary
841
831 832 832 832 834 837 838 838 839 839 839
xx
Contents
This page intentionally left blank
About the Author
Contents
xxi
About the Author Keith Jack is Director of Product Marketing at Sigma Designs, a leading supplier of Digital Media Processors that provide high-quality processing of MPEG-4, MPEG-2, MPEG-1 and Microsoft® Windows Media® 9 content. Prior to joining Sigma Designs, Mr. Jack held various marketing and chip design positions at Harris Semiconductor, Brooktree and Rockwell International. He has been involved in over 30 multimedia chips for the consumer market.
xxi
xxii
Contents
This page intentionally left blank
Contents
xxiii
What's on the CD-ROM?
What's on the CD-ROM? The included CD-ROM contains documents and tools to assist in designing, testing and evaluating various video subsystems. A fully searchable eBook version of the text in Adobe PDF format. Test images to enable the evaluation of a video subsystem. They are primarily used to test color accuracy and filter design. The sharper the transitions are without ringing, the better the filters. Since the images are computer-generated, there may be flicker along horizontal edges when viewed on an interlaced display. Unless the video signal is RF modulated, these images should pass through a system with no problems. Links to associations, licensing authorities and standards organizations. Before starting any design, and several times during the design process, verify the latest specifications are being used since they are continually updated. The list of licensing authorities should be consulted before starting any chip or system design, since several licenses are required prior to purchasing chips.
xxiii
xxiv
Contents
This page intentionally left blank
Introduction
Chapter 1: Introduction
1
Chapter 1
Introduction A few short years ago, the applications for video were somewhat confined—analog was used for broadcast and cable television, VCRs, set-top boxes, televisions and camcorders. Since then, there has been a tremendous and rapid conversion to digital video, mostly based on the MPEG-2 video compression standard. Today, in addition to the legacy DV, MPEG-1 and MPEG-2 audio and video compression standards, there are three new highperformance video compression standards. These new video codecs offer much higher video compression for a given level of video quality. • MPEG-4. This video codec typically offers a 1.5–2× improvement in compression ratio over MPEG-2. Able to address a wide variety of markets, MPEG-4 never really achieved widespread acceptance due to its complexity. Also, many simply decided to wait for the new H.264 video codec to become available.
• H.264. Also known as MPEG-4 Part 10, this video codec typically offers a 2–3× improvement in compression ratio over MPEG-2. Additional improvements in compression ratios and quality are expected as the encoders become better and use more of the available tools that H.264 offers. Learning a lesson from MPEG-4, H.264 is optimized for implementing on low-cost single-chip solutions. • SMPTE VC-9 (Microsoft® Windows Media® Video 9 or WMV9). A competitor to H.264, this video codec also typically offers a 2–3× improvement in compression ratios over MPEG-2. Again, additional improvements in compression ratios and quality are expected as the encoders become better.
1
2
Chapter 1: Introduction
Many more audio codecs are also available as a result of the interest in 6.1- and 7.1-channel audio, multi-channel lossless compression, lower bit rates for the same level of audio quality, and finally, higher bit rates for applications needing the highest audio quality at a reasonable bit rate. In addition to decoding audio, real-time high-quality audio encoding is needed for recordable DVD and digital video recorder (DVR) applications. Combining all these audio requirements mandates that any single-chip solution for the consumer market incorporate a DSP for audio processing. Equipment for the consumer has also become more sophisticated, supporting a much wider variety of content and interconnectivity. Today we have: • Networked DVD Players. In addition to playing normal CDs and DVDs, these advanced DVD players also support the playback of MPEG-4, H.264, Microsoft® Windows Media® 9 (WM9) and JPEG (for photos) content. An Ethernet or 802.11 connection enables PC-based content to be enjoyed easily on any television. Web radio and viewing of on-line movies may also be supported. • Digital Media Adapters. These small, low-cost boxes use an Ethernet or 802.11 connection to enable PC-based content to be enjoyed easily on any television. Playback of MPEG-2, MPEG-4, H.264, WM9 and JPEG content is supported.
• Digital Set-top Boxes. Cable, satellite and terrestrial set-top boxes are now including digital video recorder (DVR) capabilities, allowing viewers to enjoy content at their convenience. Many are looking at H.264 and/or WMV9 to enable system operators to offer more channels of content and reduce the chance of early obsolescence. • Advanced Digital Televisions (DTV). In addition to the tuners and decoders being incorporated inside the television, some also include an advanced DVD player, surround sound processor, wireless networking (802.11 or UWB), etc. • IP Video Set-top Boxes. Also known as “IPTV” and “video over IP”, these lowcost set-top boxes are gaining popularity in regions that have high-speed DSL and FTTH (fiber to the home) available. Many are also moving to H.264 or WMV9 to be able to offer HDTV content. • Portable Media Players. Using an internal hard disc drive (HDD), these players connect to the PC via USB or 802.11 network for downloading a wide variety of content. Playback of MPEG-2, MPEG4, H.264, WM9 and JPEG content is supported. • Mobile Video Receivers. Being incorporated into cell phones, H.264 is used to transmit a high-quality video signal. Example applications are the DMB and DVB-H standards.
Introduction
Of course, to make these advanced consumer products requires more than just supporting an audio and video codec. There is also the need to support: • Closed Captioning, Subtitles, Teletext, and V-Chip. These standards were updated to support digital broadcasts. • Advanced Video Processing. Due to the wide range of resolutions and aspect ratios for both content and displays, sophisticated high-quality scaling is usually required. Since the standard-definition (SD) and high-definition (HD) standards use different colorimetry standards, this should be corrected when viewing SD content on a HDTV or HD content on a SDTV. • Sophisticated image composition. The ability to render a sophisticated image composed of a variety of video, OSD (on-screen display), subtitle/captioning/subpicture, text and graphics elements. • ARIB and DVB over IP. The complexity of supporting IP Video is increasing, with deployments now incorporating ARIB and DVB over IP. • Digital Rights Management (DRM). The protection of content from unauthorized copying or viewing. This fourth edition of Video Demystified has been updated to reflect these changing times. Implementing “real-world” solutions is not easy, and many engineers have little knowledge or experience in this area. This book is a guide for those engineers charged with the task of understanding and implementing video features into next-generation designs.
3
This book can be used by engineers who need or desire to learn about video, VLSI design engineers working on new video products, or anyone who wants to evaluate or simply know more about video systems.
Contents The book is organized as follows: Chapter 2, an Introduction to Video, discusses the various video formats and signals, where they are used, and the differences between interlaced and progressive video. Block diagrams of DVD players and digital settop boxes are provided. Chapter 3 reviews the common Color Spaces, how they are mathematically related, and when a specific color space is used. Color spaces reviewed include RGB, YUV, YIQ, YCbCr, HSI, HSV and HLS. Considerations for converting from a non-RGB to a RGB color space and gamma correction are also discussed. Chapter 4 is a Video Signals Overview that reviews the video timing and the analog and digital representations of various video formats, including 480i, 480p, 576i, 576p, 720p, 1080i, and 1080p. Chapter 5 discusses the Analog Video Interfaces, including the analog RGB, YPbPr, SVideo and SCART interfaces for consumer and pro-video applications. Chapter 6 discusses the various Digital Video Interfaces for semiconductors, pro-video equipment and consumer equipment. It reviews the BT.601 and BT.656 semiconductor interfaces, the SDI, SDTI and HD-SDTI provideo interfaces, and the DVI, HDMI and IEEE 1394 consumer interfaces. Also reviewed are the formats for digital audio, timecode, error correction, and so on for transmission over various digital interfaces.
4
Chapter 1: Introduction
Chapter 7 covers several Digital Video Processing requirements such as 4:4:4 to 4:2:2 YCbCr, YCbCr digital filter templates, scaling, interlaced/noninterlaced conversion, frame rate conversion, alpha mixing, flicker filtering and chroma keying. Brightness, contrast, saturation, hue, and sharpness controls are also discussed. Chapter 8 provides an NTSC, PAL and SECAM Overview. The various composite analog video signal formats are reviewed, along with video test signals. VBI data discussed includes timecode, closed captioning and extended data services (XDS), widescreen signaling and teletext. In addition, PALplus, RF modulation, BTSC and Zweiton analog stereo audio and NICAM 728 digital stereo audio are reviewed. Chapter 9 covers digital techniques used for the Encoding and Decoding of NTSC and PAL color video signals. Also reviewed are various luma/chroma (Y/C) separation techniques and their trade-offs. Chapter 10 discusses the H.261 and H.263 video compression standards used for video teleconferencing. Chapter 11 discusses the Consumer DV video compression standards used by digital camcorders. Chapter 12 reviews the MPEG-1 video compression standard. Chapter 13 discusses the MPEG-2 video compression standard. Chapter 14 discusses the MPEG-4 video compression standard, including H.264. Chapter 15 discuses the ATSC Digital Television standard used in the United States. Chapter 16 discuses the OpenCable™ Digital Television standard used in the United States. Chapter 17 discuses the DVB Digital Television standard used in Europe and Asia.
Chapter 18 discuses the ISDB Digital Television standard used in Japan. Chapter 19 discusses IPTV. This technology sends compressed video over broadband networks such as Internet, DSL, FTTH (Fiber To The Home), etc. Finally, chapter 20 is a glossary of over 400 video terms which has been included for reference. If you encounter an unfamiliar term, it likely will be defined in the glossary.
Organization Addresses Many standards organizations, some of which are listed below, are involved in specifying video standards.
Advanced Television Systems Committee (ATSC) 1750 K Street NW, Suite 1200 Washington, DC 20006 Tel: (202) 872-9160 Fax: (202) 872-9161 www.atsc.org Cable Television Laboratories 858 Coal Creek Circle Louisville, CO 80027 Tel: (303) 661-9100 Fax: (303) 661-9199 www.cablelabs.com Digital Video Broadcasting (DVB) 17a Ancienne Route CH-1218 Grand Sacconnex, Geneva Switzerland Tel: +41 (0)22 717 27 14 Fax: +41 (0)22 717 27 27 www.dvb.org
Introduction
European Broadcasting Union (EBU) 17A, Ancienne Route CH-1218 Grand-Saconnex Switzerland Tel: +41 (0)22 717 2111 Fax: +41 (0)22 747 4000 www.ebu.ch Electronic Industries Alliance (EIA) 2500 Wilson Boulevard Arlington, Virginia 22201 Tel: (703) 907-7500 Fax: (703) 907-7501 www.eia.org European Telecommunications Standards Institute (ETSI) 650, route des Lucioles 06921 Sophia Antipolis Cedex France Tel: +33 (0)4 92 94 42 00 Fax: +33 (0)4 93 65 47 16 www.etsi.org International Electrotechnical Commission (IEC) 3, rue de Varembé P.O. Box 131 CH - 1211 GENEVA 20 Switzerland Tel: +41 (0)22 919 02 11 Fax: +41 (0)22 919 03 00 www.iec.ch Institute of Electrical and Electronics Engineers (IEEE) 1828 L Street, N.W., Suite 1202 Washington, D.C. 20036 Tel: (202) 785-0017 Fax: (202) 785-0835 www.ieee.org
5
International Telecommunication Union (ITU) Place des Nations CH-1211 Geneva 20 Switzerland Tel: +41 (0)22 730 5111 Fax: +41 (0)22 733 7256 www.itu.int Society of Cable Telecommunications Engineers (SCTE) 140 Philips Road Exton, PA 19341 Tel: (610) 363-6888 Fax: (610) 363-5898 www.scte.org Society of Motion Picture and Television Engineers (SMPTE) 595 West Hartsdale Avenue White Plains, NY 10607 Tel: (914) 761-1100 Fax: (914) 761-3115 www.smpte.org Video Electronics Standards Association (VESA) 920 Hillview Ct., Suite 140 Milpitas, CA 95035 Tel: (408) 957-9270 www.vesa.org
Video Demystified Web Site At the Video Demystified web site, you’ll find links to video-related newsgroups, standards, standards organizations and associations. www.video-demystified.com
6
Chapter 2: Introduction to Video
Chapter 2: Introduction to Video Chapter 2
Introduction to Video Although there are many variations and implementation techniques, video signals are just a way of transferring visual information from one point to another. The information may be from a VCR, DVD player, a channel on the local broadcast, cable television, or satellite system, the internet or one of many other sources. Invariably, the video information must be transferred from one device to another. It could be from a satellite set-top box or DVD player to a television. Or it could be from one chip to another inside the satellite set-top box or television. Although it seems simple, there are many different requirements, and therefore many different ways of doing it.
Analog vs. Digital Until a few years ago, most video equipment was designed primarily for analog video. Digital video was confined to professional applications, such as video editing.
6
The average consumer now uses digital video every day thanks to continuing falling costs. This trend has led to the development of DVD players and recorders, digital set-top boxes, digital television (DTV), portable video players and the ability to use the Internet for transferring video data.
Video Data Initially, video contained only gray-scale (also called black-and-white) information. While color broadcasts were being developed, attempts were made to transmit color video using analog RGB (red, green, blue) data. However, this technique occupied 3× more bandwidth than the current gray-scale solution, so alternate methods were developed that led to using Y, R–Y and G–Y data to represent color information. A technique was then developed to transmit this Y, R–Y and G–Y information using one signal, instead of three separate signals, and in the same bandwidth as the original gray-scale video signal. This com-
Video Timing
posite video signal is what the NTSC, PAL, and SECAM video standards are still based on today. This technique is discussed in more detail in Chapters 8 and 9. Today, even though there are many ways of representing video, they are still all related mathematically to RGB. These variations are discussed in more detail in Chapter 3. S-Video was developed for connecting consumer equipment together (it is not used for broadcast purposes). It is a set of two analog signals, one gray-scale (Y) and one that carries the analog R–Y and B–Y color information in a specific format (also called C or chroma). Once available only for S-VHS, it is now supported on most consumer video products. This is discussed in more detail in Chapter 9. Although always used by the professional video market, analog RGB video data has made a temporary come-back for connecting highend consumer equipment together. Like SVideo, it is not used for broadcast purposes. A variation of the Y, R–Y and G–Y video signals, called YPbPr, is now commonly used for connecting consumer video products together. Its primary advantage is the ability to transfer high-definition video between consumer products. Some manufacturers incorrectly label the YPbPr connectors YUV, YCbCr, or Y(B-Y)(R-Y). Chapter 5 discusses the various analog interconnect schemes in detail.
Digital Video The most common digital signals used are RGB and YCbCr. RGB is simply the digitized version of the analog RGB video signals. YCbCr is basically the digitized version of the analog YPbPr video signals, and is the format used by DVD and digital television. Chapter 6 further discusses the various digital interconnect schemes.
7
Best Connection Method There is always the question of “what is the best connection method for equipment?”. For DVD players and digital cable/satellite/ terrestrial set-top boxes, the typical order of decreasing video quality is: 1. HDMI (digital YCbCr) 2. DVI (digital RGB) 3. Analog YPbPr 4. Analog RGB 5. Analog S-Video 6. Analog Composite
Some will disagree about the order. However, most consumer products do digital video processing in the YCbCr color space. Therefore, using YCbCr as the interconnect for equipment reduces the number of color space conversions required. Color space conversion of digital signals is still preferable to D/A (digital-to-analog) conversion followed by A/D (analog-to-digital) conversion, hence the positioning of DVI above analog YPbPr. The computer industry has standardized on analog and digital RGB for connecting to the computer monitor.
Video Timing Although it looks like video is continuous motion, it is actually a series of still images, changing fast enough that it looks like continuous motion, as shown in Figure 2.1. This typically occurs 50 or 60 times per second for consumer video, and 70–90 times per second for computer displays. Special timing information, called vertical sync, is used to indicate when a new image is starting.
8
Chapter 2: Introduction to Video
IMAGE 4 IMAGE 3 IMAGE 2 IMAGE 1
TIME
Figure 2.1. Video Is Composed of a Series of Still Images. Each image is composed of individual lines of data.
Each still image is also composed of scan lines, lines of data that occur sequentially one after another down the display, as shown in Figure 2.1. Additional timing information, called horizontal sync, is used to indicate when a new scan line is starting. The vertical and horizontal sync information is usually transferred in one of three ways: 1. Separate horizontal and vertical sync signals 2. Separate composite sync signal 3. Composite sync signal embedded within the video signal
The composite sync signal is a combination of both vertical and horizontal sync. Computer and consumer equipment that uses analog RGB video usually uses technique 1 or 2. Consumer equipment that supports composite video or analog YPbPr video usually uses technique 3.
For digital video, either technique 1 is commonly used or timing code words are embedded within the digital video stream. This is discussed in Chapter 6.
Interlaced vs. Progressive Since video is a series of still images, it makes sense to simply display each full image consecutively, one after the another. This is the basic technique of progressive, or non-interlaced, displays. For progressive displays that “paint” an image on the screen, such as a CRT, each image is displayed starting at the top left corner of the display, moving to the right edge of the display. Then scanning then moves down one line, and repeats scanning left-to-right. This process is repeated until the entire screen is refreshed, as seen in Figure 2.2.
Video Resolution
In the early days of television, a technique called “interlacing” was used to reduce the amount of information sent for each image. By transferring the odd-numbered lines, followed by the even-numbered lines (as shown in Figure 2.3), the amount of information sent for each image was halved. Given this advantage of interlacing, why bother to use progressive? With interlace, each scan line is refreshed half as often as it would be if it were a progressive display. Therefore, to avoid line flicker on sharp edges due to a too-low refresh rate, the line-to-line changes are limited, essentially by vertically lowpass filtering the image. A progressive display has no limit on the line-to-line changes, so is capable of providing a higherresolution image (vertically) without flicker. Today, most broadcasts (including HDTV) are still transmitted as interlaced. Most CRTbased displays are still interlaced while LCD, plasma and computer displays are progressive.
9
Therefore, video is usually measured using “lines of resolution”. In essence, how many distinct black and white vertical lines can be seen across the display? This number is then normalized to a 1:1 display aspect ratio (dividing the number by 3/4 for a 4:3 display, or by 9/16 for a 16:9 display). Of course, this results in a lower value for widescreen (16:9) displays, which goes against intuition.
Standard Definition Standard definition video is usually defined as having 480 or 576 interlaced active scan lines, and is commonly called “480i” and “576i” respectively. For a fixed-pixel (non-CRT) consumer display with a 4:3 aspect ratio, this translates into an active resolution of 720 × 480i or 720 × 576i. For a 16:9 aspect ratio, this translates into an active resolution of 960 × 480i or 960 × 576i.
Enhanced Definition
Video Resolution Video resolution is one of those “fuzzy” things in life. It is common to see video resolutions of 720 × 480 or 1920 × 1080. However, those are just the number of horizontal samples and vertical scan lines, and do not necessarily convey the amount of useful information. For example, an analog video signal can be sampled at 13.5 MHz to generate 720 samples per line. Sampling the same signal at 27 MHz would generate 1440 samples per line. However, only the number of samples per line has changed, not the resolution of the content.
Enhanced definition video is usually defined as having 480 or 576 progressive active scan lines, and is commonly called “480p” and “576p” respectively. For a fixed-pixel (non-CRT) consumer display with a 4:3 aspect ratio, this translates into an active resolution of 720 × 480p or 720 × 576p. For a 16:9 aspect ratio, this translates into an active resolution of 960 × 480p or 960 × 576p. The difference between standard and enhanced definition is that standard definition is interlaced, while enhanced definition is progressive.
10
Chapter 2: Introduction to Video
VERTICAL
HORIZONTAL
SCANNING
SCANNING
.. .
Figure 2.2. Progressive Displays “Paint” the Lines of An Image Consecutively, One After Another.
HORIZONTAL
HORIZONTAL
VERTICAL
SCANNING
SCANNING
SCANNING
FIELD 1
FIELD 2
. . .
.. .
Figure 2.3. Interlaced Displays “Paint” First One-Half of the Image (Odd Lines), Then the Other Half (Even Lines).
Audio and Video Compression
High Definition High definition video is usually defined as having 720 progressive (720p) or 1080 interlaced (1080i) active scan lines. For a fixed-pixel (non-CRT) consumer display with a 16:9 aspect ratio, this translates into an active resolution of 1280 × 720p or 1920 × 1080i, respectively. However, HDTV displays are technically defined as being capable of displaying a minimum of 720p or 1080i active scan lines. They also must be capable of displaying 16:9 content using a minimum of 540 progressive (540p) or 810 interlaced (810i) active scan lines. This enables the manufacturing of CRT-based HDTVs with a 4:3 aspect ratio and LCD/ plasma 16:9 aspect ratio displays with resolutions of 1024 × 1024p, 1280 × 768p, 1024 × 768p, and so on, lowering costs.
Audio and Video Compression The recent advances in consumer electronics, such as digital television, DVD players and recorders, digital video recorders, and so on, were made possible due to audio and video compression based largely on MPEG-2 video with Dolby® Digital, DTS®, MPEG-1 or MPEG2 audio. New audio and video codecs, such as MPEG-4 HE-AAC, H.264 and SMPTE VC-9, offer better compression than previous codecs for the same quality. These advances are enabling new ways of distributing content (both to consumers and within the home), new consumer products (such as portable video players and mobile video/cell phones) and more cable/satellite channels.
11
Application Block Diagrams Looking at a few simplified block diagrams helps envision how video flows through its various operations.
DVD Players Figure 2.4 is a simplified block diagram for a basic DVD-Video player, showing the common blocks. Today, all of this is on a single lowcost chip. In addition to playing DVDs (which are based on MPEG-2 video compression), DVD players are now expected to handle MP3 and WMA audio, MPEG-4 video (for DivX Video), JPEG images, and so on. Special playback modes such as slow/fast forward/reverse at various speeds are also expected. Support for DVD-Audio and SACD is also increasing. A recent enhancement to DVD players is the ability to connect to a home network for playing content (music, video, pictures, etc.) residing on the PC. These “networked DVD players” may also include the ability to play movies from the Internet and download content onto an internal hard disc drive (HDD) for later viewing. Support for playing audio, video and pictures from a variety of flash-memory cards is also growing. In an attempt to “look different” to quickly grab buyers attention, some DVD player manufacturers “tweak” the video frequency response. Since this “feature” is usually irritating over the long term, it should be defeated or properly adjusted. For the “film look” many video enthusiasts strive for, the frequency response should be as flat as possible. Another issue is the output levels of the analog video signals. Although it is easy to generate very accurate video levels, they vary con-
12
Chapter 2: Introduction to Video
CLOSED CAPTIONING, TELETEXT, WIDESCREEN VBI DATA
SCALING
VIDEO
FROM READ ELECTRONICS
CSS
DECOMPRESS
DESCRAMBLE
(MPEG 2)
--------------
BRIGHTNESS CONTRAST
GRAPHICS
NTSC / PAL
HUE
OVERLAY
VIDEO ENCODE
SATURATION
NTSC / PAL, S-VIDEO
RGB / YPBPR HDMI
SHARPNESS
PROGRAM STREAM DEMUX
AUDIO
STEREO AUDIO
AUDIO L
DAC
AUDIO R
DECOMPRESS (DOLBY DIGITAL AND DTS)
DIGITAL AUDIO INTERFACE
5.1 DIGITAL AUDIO
CPU
IR INPUT
Figure 2.4. Simplified Block Diagram of a Basic DVD Player.
siderably. Reviews are now pointing out this issue since switching between sources may mean changing brightness or black levels, defeating any television calibration or personal adjustments that may have been done by the user.
Digital Television Set-top Boxes The digital television standards fall into six major categories: ATSC (Advanced Television Systems Committee) DVB (Digital Video Broadcast)
Digital Media Adapters Digital media adapters connect to a home network for playing content (music, video, pictures, and so on) residing on the PC. These small, low-cost boxes enable PC-based content to be enjoyed on any or all televisions in the home. Many support optional wireless networking, simplifying installation. Except for DVD playback, they have capabilities similar to networked DVD players. Figure 2.5 is a simplified block diagram for a basic digital media adapter, showing the common blocks. Today, all of this is on a single lowcost chip.
ARIB (Association of Radio Industries and Businesses) Open digital cable standards, such as OpenCable Proprietary digital cable standards Proprietary digital satellite standards
Currently based on MPEG-2 video compression, with Dolby® Digital or MPEG audio compression, work is progressing on supporting the new advanced audio and video codecs, such as HE-AAC, H.264 and VC-9.
Application Block Diagrams
13
CLOSED CAPTIONING, TELETEXT, WIDESCREEN VBI DATA
SCALING BRIGHTNESS VIDEO
CONTRAST
GRAPHICS
NTSC / PAL
DECOMPRESS
HUE
OVERLAY
VIDEO ENCODE
SATURATION ETHERNET NETWORK
NTSC / PAL, S-VIDEO RGB / YPBPR HDMI
SHARPNESS
DEMUX
STEREO AUDIO
AUDIO L
DAC
AUDIO R
AUDIO DECOMPRESS DIGITAL AUDIO INTERFACE
IR
5.1 DIGITAL AUDIO
CPU
INPUT
Figure 2.5. Simplified Block Diagram of a Digital Media Adapter.
Figure 2.6 is a simplified block diagram for a digital television set-top box, showing the common audio and video processing blocks. It is used to receive digital television broadcasts, from either terrestrial (over-the-air), cable, or satellite. A digital television may include this circuitry inside the television.
Many set-top boxes now include two tuners and digital video recorder (DVR) capability. This enables recording one program onto an internal HDD while watching another. Two tuners are also common in digital television receivers to support a picture-in-picture (PIP) feature.
14 Chapter 2: Introduction to Video
Figure 2.6. Simplified Block Diagram of a Digital Television Set-top Box.
CLOSED CAPTIONING, TELETEXT, WIDESCREEN VBI DATA
SCALING
QAM / VSB / RF INPUT
COFDM DEMOD TUNER
AND FEC
CHANNEL
VIDEO DECOMPRESS
DESCRAMBLE
(MPEG 2)
BRIGHTNESS CONTRAST
GRAPHICS
NTSC / PAL
HUE
OVERLAY
VIDEO ENCODE
SATURATION
--------------
NTSC / PAL, S-VIDEO RGB / YPBPR VIDEO HDMI
SHARPNESS
TRANSPORT STREAM DEMUX
AUDIO
STEREO AUDIO
AUDIO L
DAC
AUDIO R
DECOMPRESS (DOLBY DIGITAL DTS, MPEG)
DIGITAL AUDIO INTERFACE
NTSC / PAL VIDEO DECODE NTSC / PAL AUDIO DECODE
IR INPUT
CPU
5.1 DIGITAL AUDIO
RGB Color Space
Chapter 3: Color Spaces
15
Chapter 3
Color Spaces A color space is a mathematical representation of a set of colors. The three most popular color models are RGB (used in computer graphics); YIQ, YUV, or YCbCr (used in video systems); and CMYK (used in color printing). However, none of these color spaces are directly related to the intuitive notions of hue, saturation, and brightness. This resulted in the temporary pursuit of other models, such as HSI and HSV, to simplify programming, processing, and end-user manipulation. All of the color spaces can be derived from the RGB information supplied by devices such as cameras and scanners.
RGB Color Space The red, green, and blue (RGB) color space is widely used throughout computer graphics. Red, green, and blue are three primary additive colors (individual components are added together to form a desired color) and are represented by a three-dimensional, Cartesian coordinate system (Figure 3.1). The indicated diagonal of the cube, with equal amounts of each primary component, represents various gray levels. Table 3.1 contains the RGB values for 100% amplitude, 100% saturated color bars, a common video test signal.
CYAN
BLUE
MAGENTA
WHITE
BLACK
RED
GREEN
YELLOW
Figure 3.1. The RGB Color Cube.
15
White
Yellow
Cyan
Green
Magenta
Red
Blue
Black
Chapter 3: Color Spaces
Nominal Range
16
R
0 to 255
255
255
0
0
255
255
0
0
G
0 to 255
255
255
255
255
0
0
0
0
B
0 to 255
255
0
255
0
255
0
255
0
Table 3.1. 100% RGB Color Bars.
The RGB color space is the most prevalent choice for computer graphics because color displays use red, green and blue to create the desired color. Therefore, the choice of the RGB color space simplifies the architecture and design of the system. Also, a system that is designed using the RGB color space can take advantage of a large number of existing software routines, since this color space has been around for a number of years. However, RGB is not very efficient when dealing with “real-world” images. All three RGB components need to be of equal bandwidth to generate any color within the RGB color cube. The result of this is a frame buffer that has the same pixel depth and display resolution for each RGB component. Also, processing an image in the RGB color space is usually not the most efficient method. For example, to modify the intensity or color of a given pixel, the three RGB values must be read from the frame buffer, the intensity or color calculated, the desired modifications performed, and the new RGB values calculated and written back to the frame buffer. If the system had access to an image stored directly in the intensity and color format, some processing steps would be faster. For these and other reasons, many video standards use luma and two color difference signals. The most common are the YUV, YIQ,
and YCbCr color spaces. Although all are related, there are some differences.
YUV Color Space The YUV color space is used by the PAL (Phase Alternation Line), NTSC (National Television System Committee), and SECAM (Sequentiel Couleur Avec Mémoire or Sequential Color with Memory) composite color video standards. The black-and-white system used only luma (Y) information; color information (U and V) was added in such a way that a black-and-white receiver would still display a normal black-and-white picture. Color receivers decoded the additional color information to display a color picture. The basic equations to convert between gamma-corrected RGB (notated as R´G´B´ and discussed later in this chapter) and YUV are: Y = 0.299R´ + 0.587G´ + 0.114B´ U = – 0.147R´ – 0.289G´ + 0.436B´ = 0.492 (B´ – Y) V = 0.615R´ – 0.515G´ – 0.100B´ = 0.877(R´ – Y)
YIQ Color Space
R´ = Y + 1.140V
Y = 0.299R´ + 0.587G´ + 0.114B´
G´ = Y – 0.395U – 0.581V
I = 0.596R´ – 0.275G´ – 0.321B´ = Vcos 33° – Usin 33° = 0.736(R´ – Y) – 0.268(B´ – Y)
B´ = Y + 2.032U
For digital R´G´B´ values with a range of 0– 255, Y has a range of 0–255, U a range of 0 to ±112, and V a range of 0 to ±157. These equations are usually scaled to simplify the implementation in an actual NTSC or PAL digital encoder or decoder. Note that for digital data, 8-bit YUV and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems. If the full range of (B´ – Y) and (R´ – Y) had been used, the composite NTSC and PAL levels would have exceeded what the (then current) black-and-white television transmitters and receivers were capable of supporting. Experimentation determined that modulated subcarrier excursions of 20% of the luma (Y) signal excursion could be permitted above white and below black. The scaling factors were then selected so that the maximum level of 75% amplitude, 100% saturation yellow and cyan color bars would be at the white level (100 IRE).
YIQ Color Space The YIQ color space, further discussed in Chapter 8, is derived from the YUV color space and is optionally used by the NTSC composite color video standard. (The “I” stands for “inphase” and the “Q” for “quadrature,” which is the modulation method used to transmit the color information.) The basic equations to convert between R´G´B´ and YIQ are:
17
Q = 0.212R´ – 0.523G´ + 0.311B´ = Vsin 33° + Ucos 33° = 0.478(R´ – Y) + 0.413(B´ – Y)
or, using matrix notation: I = 0 1 cos ( 33 ) sin ( 33 ) U Q 1 0 – sin ( 33 ) cos ( 33 ) V R´ = Y + 0.956I + 0.621Q G´ = Y – 0.272I – 0.647Q B´ = Y – 1.107I + 1.704Q
For digital R´G´B´ values with a range of 0– 255, Y has a range of 0–255, I has a range of 0 to ±152, and Q has a range of 0 to ±134. I and Q are obtained by rotating the U and V axes 33°. These equations are usually scaled to simplify the implementation in an actual NTSC digital encoder or decoder. Note that for digital data, 8-bit YIQ and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems.
YCbCr Color Space The YCbCr color space was developed as part of ITU-R BT.601 during the development of a world-wide digital component video standard (discussed in Chapter 4). YCbCr is a scaled and offset version of the YUV color space. Y is defined to have a nominal 8-bit
18
Chapter 3: Color Spaces
range of 16–235; Cb and Cr are defined to have a nominal range of 16–240. There are several YCbCr sampling formats, such as 4:4:4, 4:2:2, 4:1:1, and 4:2:0 that are also described.
When performing YCbCr to R´G´B´ conversion, the resulting R´G´B´ values have a nominal range of 16–235, with possible occasional excursions into the 0–15 and 236–255 values. This is due to Y and CbCr occasionally going outside the 16–235 and 16–240 ranges, respectively, due to video processing and noise. Note that 8-bit YCbCr and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems. Table 3.2 lists the YCbCr values for 75% amplitude, 100% saturated color bars, a common video test signal.
RGB - YCbCr Equations: SDTV The basic equations to convert between 8bit digital R´G´B´ data with a 16–235 nominal range (Studio RGB) and YCbCr are: Y601 = 0.299R´ + 0.587G´ + 0.114B´ Cb = –0.172R´ – 0.339G´ + 0.511B´ + 128 Cr = 0.511R´ – 0.428G´ – 0.083B´ + 128
Computer Systems Considerations If the R´G´B´ data has a range of 0–255, as is commonly found in computer systems, the following equations may be more convenient to use:
R´ = Y601 + 1.371(Cr – 128) G´ = Y601 – 0.698(Cr – 128) – 0.336(Cb – 128) B´ = Y601 + 1.732(Cb – 128)
Y601 = 0.257R´ + 0.504G´ + 0.098B´ + 16 Cb = –0.148R´ – 0.291G´ + 0.439B´ + 128
Red
Blue
Black
16 to 240
Magenta
16 to 240
Cr
Green
Cb
180
162
131
112
84
65
35
16
128
44
156
72
184
100
212
128
128
142
44
58
198
212
114
128
Cyan
16 to 235
Yellow
Y
White
Nominal Range
Cr = 0.439R´ – 0.368G´ – 0.071B´ + 128
SDTV
HDTV Y
16 to 235
180
168
145
133
63
51
28
16
Cb
16 to 240
128
44
147
63
193
109
212
128
Cr
16 to 240
128
136
44
52
204
212
120
128
Table 3.2. 75% YCbCr Color Bars.
YCbCr Color Space
R´ = 1.164(Y601 – 16) + 1.596(Cr – 128) G´ = 1.164(Y601 – 16) – 0.813(Cr – 128) – 0.391(Cb – 128) B´ = 1.164(Y601 – 16) + 2.018(Cb – 128)
Note that 8-bit YCbCr and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems.
RGB - YCbCr Equations: HDTV The basic equations to convert between 8bit digital R´G´B´ data with a 16–235 nominal range (Studio RGB) and YCbCr are: Y709 = 0.213R´ + 0.715G´ + 0.072B´ Cb = –0.117R´ – 0.394G´ + 0.511B´ + 128 Cr = 0.511R´ – 0.464G´ – 0.047B´ + 128 R´ = Y709 + 1.540(Cr – 128) G´ = Y709 – 0.459(Cr – 128) – 0.183(Cb – 128) B´ = Y709 + 1.816(Cb – 128)
When performing YCbCr to R´G´B´ conversion, the resulting R´G´B´ values have a nominal range of 16–235, with possible occasional excursions into the 0–15 and 236–255 values. This is due to Y and CbCr occasionally going outside the 16–235 and 16–240 ranges, respectively, due to video processing and noise. Note that 8-bit YCbCr and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems. Table 3.2 lists the YCbCr values for 75% amplitude, 100% saturated color bars, a common video test signal.
19
Computer Systems Considerations If the R´G´B´ data has a range of 0–255, as is commonly found in computer systems, the following equations may be more convenient to use: Y709 = 0.183R´ + 0.614G´ + 0.062B´ + 16 Cb = –0.101R´ – 0.338G´ + 0.439B´ + 128 Cr = 0.439R´ – 0.399G´ – 0.040B´ + 128 R´ = 1.164(Y709 – 16) + 1.793(Cr – 128) G´ = 1.164(Y709 – 16) – 0.534(Cr – 128) – 0.213(Cb – 128) B´ = 1.164(Y709 – 16) + 2.115(Cb – 128)
Note that 8-bit YCbCr and R´G´B´ data should be saturated at the 0 and 255 levels to avoid underflow and overflow wrap-around problems.
4:4:4 YCbCr Format Figure 3.2 illustrates the positioning of YCbCr samples for the 4:4:4 format. Each sample has a Y, a Cb and a Cr value. Each sample is typically 8 bits (consumer applications) or 10 bits (pro-video applications) per component. Each sample therefore requires 24 bits (or 30 bits for pro-video applications).
4:2:2 YCbCr Format Figure 3.3 illustrates the positioning of YCbCr samples for the 4:2:2 format. For every two horizontal Y samples, there is one Cb and Cr sample. Each sample is typically 8 bits (consumer applications) or 10 bits (pro-video applications) per component. Each sample therefore requires 16 bits (or 20 bits for provideo applications), usually formatted as shown in Figure 3.4.
20
Chapter 3: Color Spaces
To display 4:2:2 YCbCr data, it is first converted to 4:4:4 YCbCr data, using interpolation to generate the missing Cb and Cr samples.
4:1:1 YCbCr Format Figure 3.5 illustrates the positioning of YCbCr samples for the 4:1:1 format (also known as YUV12), used in some consumer video and DV video compression applications. For every four horizontal Y samples, there is one Cb and Cr value. Each component is typically 8 bits. Each sample therefore requires 12 bits, usually formatted as shown in Figure 3.6. To display 4:1:1 YCbCr data, it is first converted to 4:4:4 YCbCr data, using interpolation to generate the missing Cb and Cr samples.
ACTIVE LINE NUMBER
X = FIELD 1 (576i FIELD 2) [ X ] = FIELD 2 (576i FIELD 1)
4:2:0 YCbCr Format Rather than the horizontal-only 2:1 reduction of Cb and Cr used by 4:2:2, 4:2:0 YCbCr implements a 2:1 reduction of Cb and Cr in both the vertical and horizontal directions. It is commonly used for video compression. As shown in Figures 3.7 through 3.11, there are several 4:2:0 sampling formats. Table 3.3 lists the YCbCr formats for various DV applications. To display 4:2:0 YCbCr data, it is first converted to 4:4:4 YCbCr data, using interpolation to generate the new Cb and Cr samples. Note that some MPEG decoders do not properly convert the 4:2:0 YCbCr data to the 4:4:4 format, resulting in a “chroma bug.”
ACTIVE LINE NUMBER
1
1
[1]
[1]
2
2
[2]
[2]
3
3
X = FIELD 1 (576i FIELD 2) [ X ] = FIELD 2 (576i FIELD 1)
CB, CR SAMPLE
CB, CR SAMPLE
Y SAMPLE
Y SAMPLE
Figure 3.2. 4:4:4 Co-Sited Sampling. The sampling positions on the active scan lines of an interlaced picture.
Figure 3.3. 4:2:2 Co-Sited Sampling. The sampling positions on the active scan lines of an interlaced picture.
YCbCr Color Space
SAMPLE 0
SAMPLE SAMPLE SAMPLE SAMPLE SAMPLE 1 2 3 4 5
Y7 - 0 Y6 - 0 Y5 - 0 Y4 - 0 Y3 - 0 Y2 - 0 Y1 - 0 Y0 - 0
Y7 - 1 Y6 - 1 Y5 - 1 Y4 - 1 Y3 - 1 Y2 - 1 Y1 - 1 Y0 - 1
Y7 - 2 Y6 - 2 Y5 - 2 Y4 - 2 Y3 - 2 Y2 - 2 Y1 - 2 Y0 - 2
Y7 - 3 Y6 - 3 Y5 - 3 Y4 - 3 Y3 - 3 Y2 - 3 Y1 - 3 Y0 - 3
Y7 - 4 Y6 - 4 Y5 - 4 Y4 - 4 Y3 - 4 Y2 - 4 Y1 - 4 Y0 - 4
Y7 - 5 Y6 - 5 Y5 - 5 Y4 - 5 Y3 - 5 Y2 - 5 Y1 - 5 Y0 - 5
CB7 - 0 CB6 - 0 CB5 - 0 CB4 - 0 CB3 - 0 CB2 - 0 CB1 - 0 CB0 - 0
CR7 - 0 CR6 - 0 CR5 - 0 CR4 - 0 CR3 - 0 CR2 - 0 CR1 - 0 CR0 - 0
CB7 - 2 CB6 - 2 CB5 - 2 CB4 - 2 CB3 - 2 CB2 - 2 CB1 - 2 CB0 - 2
CR7 - 2 CR6 - 2 CR5 - 2 CR4 - 2 CR3 - 2 CR2 - 2 CR1 - 2 CR0 - 2
CB7 - 4 CB6 - 4 CB5 - 4 CB4 - 4 CB3 - 4 CB2 - 4 CB1 - 4 CB0 - 4
CR7 - 4 CR6 - 4 CR5 - 4 CR4 - 4 CR3 - 4 CR2 - 4 CR1 - 4 CR0 - 4
-0 -1 -2 -3 -4
= = = = =
21
SAMPLE SAMPLE SAMPLE SAMPLE SAMPLE
0 1 2 3 4
16 BITS PER SAMPLE
DATA DATA DATA DATA DATA
Figure 3.4. 4:2:2 Frame Buffer Formatting.
SAMPLE SAMPLE SAMPLE SAMPLE SAMPLE SAMPLE 0 1 2 3 4 5 ACTIVE LINE NUMBER
X = FIELD 1 (576i FIELD 2) [ X ] = FIELD 2 (576i FIELD 1)
1 [1]
Y7 - 0 Y6 - 0 Y5 - 0 Y4 - 0 Y3 - 0 Y2 - 0 Y1 - 0 Y0 - 0
Y7 - 1 Y6 - 1 Y5 - 1 Y4 - 1 Y3 - 1 Y2 - 1 Y1 - 1 Y0 - 1
Y7 - 2 Y6 - 2 Y5 - 2 Y4 - 2 Y3 - 2 Y2 - 2 Y1 - 2 Y0 - 2
Y7 - 3 Y6 - 3 Y5 - 3 Y4 - 3 Y3 - 3 Y2 - 3 Y1 - 3 Y0 - 3
Y7 - 4 Y6 - 4 Y5 - 4 Y4 - 4 Y3 - 4 Y2 - 4 Y1 - 4 Y0 - 4
Y7 - 5 Y6 - 5 Y5 - 5 Y4 - 5 Y3 - 5 Y2 - 5 Y1 - 5 Y0 - 5
CB7 - 0 CB6 - 0 CR7 - 0 CR6 - 0
CB5 - 0 CB4 - 0 CR5 - 0 CR4 - 0
CB3 - 0 CB2 - 0 CR3 - 0 CR2 - 0
CB1 - 0 CB0 - 0 CR1 - 0 CR0 - 0
CB7 - 4 CB6 - 4 CR7 - 4 CR6 - 4
CB5 - 4 CB4 - 4 CR5 - 4 CR4 - 4
12 BITS PER SAMPLE
2 [2] 3
CB, CR SAMPLE Y SAMPLE
Figure 3.5. 4:1:1 Co-Sited Sampling. The sampling positions on the active scan lines of an interlaced picture.
-0 -1 -2 -3 -4
= = = = =
SAMPLE SAMPLE SAMPLE SAMPLE SAMPLE
0 1 2 3 4
DATA DATA DATA DATA DATA
Figure 3.6. 4:1:1 Frame Buffer Formatting.
Chapter 3: Color Spaces
ACTIVE LINE NUMBER
ACTIVE LINE NUMBER
1
1
2
2
3
3
4
4
5
5
CALCULATED CB, CR SAMPLE
CALCULATED CB, CR SAMPLE
Y SAMPLE
Y SAMPLE
4:1:1 Co-Sited 4:2:0 4:2:0 Co-Sited
×
×
D-9 HD
×
DVCPRO HD
×
D-9 (Digital S)
4:2:2 Co-Sited
Digital Betacam
4:4:4 Co-Sited
100 Mbps DV
DVCPRO 50
D-7 (DVCPRO)
50 Mbps DV 576-Line DVCAM
480-Line DVCAM
576-Line DV
YCbCr Format
480-Line DV
25 Mbps DV
×
×
×
×
×
×
Table 3.3. YCbCr Formats for Various DV Applications.
×
H.261, H.263
Figure 3.8. 4:2:0 Sampling for MPEG-2, MPEG-4 Part 2 and H.264. The sampling positions on the active scan lines of a progressive or noninterlaced picture.
MPEG-2, -4 Part 2, H.264
Figure 3.7. 4:2:0 Sampling for H.261, H.263, and MPEG-1. The sampling positions on the active scan lines of a progressive or noninterlaced picture.
MPEG-1
22
× ×
×
×
YCbCr Color Space ACTIVE LINE NUMBER
FIELD N
23
FIELD N + 1
1 [1] 2 [2] 3 [3] 4 [4]
CALCULATED CB, CR SAMPLE Y SAMPLE
Figure 3.9. 4:2:0 Sampling for MPEG-2, MPEG-4 Part 2 and H.264. The sampling positions on the active scan lines of an interlaced picture (top_field_first = 1).
ACTIVE LINE NUMBER
FIELD N
FIELD N + 1
1 [1] 2 [2] 3 [3] 4 [4]
CALCULATED CB, CR SAMPLE Y SAMPLE
Figure 3.10. 4:2:0 Sampling for MPEG-2, MPEG-4 Part 2 and H.264. The sampling positions on the active scan lines of an interlaced picture (top_field_first = 0).
24
Chapter 3: Color Spaces
ACTIVE LINE NUMBER
FIELD N + 1
FIELD N
1 [1] 2 [2] 3 [3] 4 [4]
CR SAMPLE CB SAMPLE Y SAMPLE
Figure 3.11. 4:2:0 Co-Sited Sampling for 576i DV and DVCAM. The sampling positions on the active scan lines of an interlaced picture.
PhotoYCC Color Space PhotoYCC (a trademark of Eastman Kodak Company) was developed to encode Photo CD image data. The goal was to develop a display-device-independent color space. For maximum video display efficiency, the color space is based upon ITU-R BT.601 and BT.709. The encoding process (RGB to PhotoYCC) assumes CIE Standard Illuminant D65 and that the spectral sensitivities of the image capture system are proportional to the color-matching functions of the BT.709 reference primaries. The RGB values, unlike those for a computer graphics system, may be negative. PhotoYCC includes colors outside the BT.709 color gamut; these are encoded using negative values.
RGB to PhotoYCC Linear RGB data (normalized to have values of 0 to 1) is nonlinearly transformed to PhotoYCC as follows: for R, G, B ≥ 0.018 R´ = 1.099 R0.45 – 0.099 G´ = 1.099 G0.45 – 0.099 B´ = 1.099 B0.45 – 0.099
for –0.018 < R, G, B < 0.018 R´ = 4.5 R G´ = 4.5 G B´ = 4.5 B
HSI, HLS and HSV Color Spaces for R, G, B ≤ –0.018
R´ = 0.981Y + 1.315(C2 – 137)
0.45
– 0.099
G´ = 0.981Y – 0.311(C1 – 156) – 0.669(C2 – 137)
|G|0.45
– 0.099
B´ = 0.981Y + 1.601 (C1 – 156)
0.45
– 0.099
R´ = – 1.099 |R| G´ = – 1.099
25
B´ = – 1.099 |B|
From R´G´B´ with a 0–255 range, a luma and two chrominance signals (C1 and C2) are generated: Y = 0.213R´ + 0.419G´ + 0.081B´ C1 = – 0.131R´ – 0.256G´ + 0.387B´ + 156 C2 = 0.373R´ – 0.312G´ – 0.061B´ + 137
As an example, a 20% gray value (R, G, and B = 0.2) would be recorded on the PhotoCD disc using the following values: Y = 79 C1 = 156 C2 = 137
PhotoYCC to RGB Since PhotoYCC attempts to preserve the dynamic range of film, decoding PhotoYCC images requires the selection of a color space and range appropriate for the output device. Thus, the decoding equations are not always the exact inverse of the encoding equations. The following equations are suitable for generating RGB values for driving a CRT display, and assume a unity relationship between the luma in the encoded image and the displayed image.
The R´G´B´ values should be saturated to a range of 0 to 255. The equations above assume the display uses phosphor chromaticities that are the same as the BT.709 reference primaries, and that the video signal luma (V) and the display luminance (L) have the relationship: for V ≥ 0.0812 L = ((V + 0.099) / 1.099)1/0.45
for V < 0.0812 L = V / 4.5
HSI, HLS and HSV Color Spaces The HSI (hue, saturation, intensity) and HSV (hue, saturation, value) color spaces were developed to be more “intuitive” in manipulating color and were designed to approximate the way humans perceive and interpret color. They were developed when colors had to be specified manually, and are rarely used now that users can select colors visually or specify Pantone colors. These color spaces are discussed for “historic” interest. HLS (hue, lightness, saturation) is similar to HSI; the term lightness is used rather than intensity. The difference between HSI and HSV is the computation of the brightness component (I or V), which determines the distribution and dynamic range of both the brightness (I or V) and saturation (S). The HSI color space is best
26
Chapter 3: Color Spaces
for traditional image processing functions such as convolution, equalization, histograms, and so on, which operate by manipulation of the brightness values since I is equally dependent on R, G, and B. The HSV color space is preferred for manipulation of hue and saturation (to shift colors or adjust the amount of color) since it yields a greater dynamic range of saturation. Figure 3.12 illustrates the single hexcone HSV color model. The top of the hexcone corresponds to V = 1, or the maximum intensity colors. The point at the base of the hexcone is black and here V = 0. Complementary colors are 180° opposite one another as measured by H, the angle around the vertical axis (V), with red at 0°. The value of S is a ratio, ranging from 0 on the center line vertical axis (V) to 1 on the sides of the hexcone. Any value of S between 0 and 1 may be associated with the point V = 0. The point S = 0, V = 1 is white. Intermediate values of V for S = 0 are the grays. Note that when S = 0, the value of H is irrelevant. From an artist’s viewpoint, any color with V = 1, S = 1 is a pure pigment (whose color is defined by H). Adding white corresponds to decreasing S (without changing V); adding black corresponds to decreasing V (without changing S). Tones are created by decreasing both S and V. Table 3.4 lists the 75% amplitude, 100% saturated HSV color bars. Figure 3.13 illustrates the double hexcone HSI color model. The top of the hexcone corresponds to I = 1, or white. The point at the base of the hexcone is black and here I = 0. Complementary colors are 180° opposite one another as measured by H, the angle around the vertical axis (I), with red at 0° (for consistency with the HSV model, we have changed from the Tektronix convention of blue at 0°). The value of S ranges from 0 on the vertical axis (I) to 1 on the surfaces of the hexcone. The grays all have S = 0, but maximum saturation of hues is
at S = 1, I = 0.5. Table 3.5 lists the 75% amplitude, 100% saturated HSI color bars.
Chromaticity Diagram The color gamut perceived by a person with normal vision (the 1931 CIE Standard Observer) is shown in Figure 3.14. The diagram and underlying mathematics were updated in 1960 and 1976; however, the NTSC television system is based on the 1931 specifications. Color perception was measured by viewing combinations of the three standard CIE (International Commission on Illumination or Commission Internationale de I’Eclairage) primary colors: red with a 700-nm wavelength, green at 546.1 nm, and blue at 435.8 nm. These primary colors, and the other spectrally pure colors resulting from mixing of the primary colors, are located along the curved outer boundary line (called the spectrum locus), shown in Figure 3.14. The ends of the spectrum locus (at red and blue) are connected by a straight line that represents the purples, which are combinations of red and blue. The area within this closed boundary contains all the colors that can be generated by mixing light of different colors. The closer a color is to the boundary, the more saturated it is. Colors within the boundary are perceived as becoming more pastel as the center of the diagram (white) is approached. Each point on the diagram, representing a unique color, may be identified by its x and y coordinates. In the CIE system, the intensities of red, green, and blue are transformed into what are called the tristimulus values, which are represented by the capital letters X, Y, and Z. These values represent the relative quantities of the primary colors.
Chromaticity Diagram
V
GREEN 120˚
CYAN 180˚
WHITE
YELLOW 60˚
RED 0˚
1.0
BLUE 240˚
MAGENTA 300˚
H 0.0
S BLACK
Cyan
Green
Magenta
–
60°
180°
120°
300°
S
0 to 1
0
1
1
1
1
V
0 to 1
0.75
0.75
0.75
0.75
0.75
Table 3.4. 75% HSV Color Bars.
Black
Yellow
0° to 360°
Blue
White
H
Red
Nominal Range
Figure 3.12. Single Hexcone HSV Color Model.
0°
240°
–
1
1
0
0.75
0.75
0
27
Chapter 3: Color Spaces
I WHITE 1.0
GREEN 120˚
YELLOW 60˚
CYAN 180˚
RED 0˚
BLUE 240˚
MAGENTA 300˚
H S BLACK 0.0
Yellow
Cyan
Green
Magenta
Red
Blue
H
0° to 360°
–
60°
180°
120°
300°
0°
240°
–
S
0 to 1
0
1
1
1
1
1
1
0
I
0 to 1
0.75
0.375
0.375
0.375
0.375
0.375
0.375
0
Black
White
Figure 3.13. Double Hexcone HSI Color Model. For consistency with the HSV model, we have changed from the Tektronix convention of blue at 0° and depict the model as a double hexcone rather than as a double cone.
Nominal Range
28
Table 3.5. 75% HSI Color Bars. For consistency with the HSV model, we have changed from the Tektronix convention of blue at 0°.
Chromaticity Diagram
The coordinate axes of Figure 3.14 are derived from the tristimulus values: x = X/(X + Y + Z) = red/(red + green + blue) y = Y/(X + Y + Z) = green/(red + green + blue) z = Z/(X + Y + Z) = blue/(red + green + blue)
The coordinates x, y, and z are called chromaticity coordinates, and they always add up to 1. As a result, z can always be expressed in terms of x and y, which means that only x and y are required to specify any color, and the diagram can be two-dimensional. Typically, a source or display specifies three (x, y) coordinates to define the three primary colors it uses. The triangle formed by the three (x, y) coordinates encloses the gamut of colors that the source or display can reproduce. This is shown in Figure 3.15, which compares the color gamuts of NTSC, PAL, and typical inks and dyes. Note that no set of three colors can generate all possible colors, which is why television pictures are never completely accurate. In addition, a source or display usually specifies the (x, y) coordinate of the white color used, since pure white is not usually captured or reproduced. White is defined as the color captured or produced when all three primary signals are equal, and it has a subtle shade of color to it. Note that luminance, or brightness information, is not included in the standard CIE 1931 chromaticity diagram, but is an axis that is orthogonal to the (x, y) plane. The lighter a color is, the more restricted the chromaticity range is.
29
The chromaticities and reference white (CIE illuminate C) for the 1953 NTSC standard are: R:
xr = 0.67
yr = 0.33
G:
xg = 0.21
yg = 0.71
B:
xb = 0.14
yb = 0.08
white:
xw = 0.3101
yw = 0.3162
Modern NTSC, 480i and 480p video systems use a different set of RGB phosphors, resulting in slightly different chromaticities of the RGB primaries and reference white (CIE illuminate D65): R:
xr = 0.630
yr = 0.340
G:
xg = 0.310
yg = 0.595
B:
xb = 0.155
yb = 0.070
white:
xw = 0.3127
yw = 0.3290
The chromaticities and reference white (CIE illuminate D65) for PAL, SECAM, 576i and 576p video systems are: R:
xr = 0.64
yr = 0.33
G:
xg = 0.29
yg = 0.60
B:
xb = 0.15
yb = 0.06
white:
xw = 0.3127
yw = 0.3290
The chromaticities and reference white (CIE illuminate D65) for HDTV are based on BT.709: R:
xr = 0.64
yr = 0.33
G:
xg = 0.30
yg = 0.60
B:
xb = 0.15
yb = 0.06
white:
xw = 0.3127
yw = 0.3290
30
Chapter 3: Color Spaces
Since different chromaticity and reference white values are used for various video standards, minor color errors may occur when the source and display values do not match; for example, displaying a 480i or 480p program on a HDTV, or displaying a HDTV program on a NTSC television. These minor color errors can easily be corrected at the display by using a 3 × 3 matrix multiplier, as discussed in Chapter 7.
Non-RGB Color Space Considerations When processing information in a nonRGB color space (such as YIQ, YUV, or YCbCr), care must be taken that combinations of values are not created that result in the generation of invalid RGB colors. The term invalid refers to RGB components outside the normalized RGB limits of (1, 1, 1).
y
y
1.0
1.0
0.9
0.9 520
520 0.8
0.8
GREEN
540
GREEN
540
0.7
0.7
1953 NTSC COLOR GAMUT NTSC / PAL / SECAM / HDTV COLOR GAMUT
560
560 INK / DYE COLOR GAMUT
0.6
0.6
500
500 YELLOW
0.5
0.5
580
580
ORANGE 0.4
0.4
600
600
WHITE CYAN
0.3
PINK
0.3 RED
RED
780 NM
780 NM
0.2
0.2 480
BLUE
480
PURPLE
0.1
0.1 BLUE
380 0.0
0.1
0.2
0.3
380
0.0
0.0 0.4
0.5
0.6
0.7
0.8
x
Figure 3.14. CIE 1931 Chromaticity Diagram Showing Various Color Regions.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
x
Figure 3.15. CIE 1931 Chromaticity Diagram Showing Various Color Gamuts.
Non-RGB Color Space Considerations
31
ALL POSSIBLE YCBCR VALUES
Y = 255, CB = CR = 128
W
Y 255
C G R
Y
M
255
0
B
BK
CR
CB
255
YCBCR VALID COLOR BLOCK
R = RED G = GREEN B = BLUE Y = YELLOW C = CYAN M = MAGENTA W = WHITE BK = BLACK
Figure 3.16. RGB Limits Transformed into 3-D YCbCr Space.
For example, given that RGB has a normalized value of (1, 1, 1), the resulting YCbCr value is (235, 128, 128). If Cb and Cr are manipulated to generate a YCbCr value of (235, 64, 73), the corresponding RGB normalized value becomes (0.6, 1.29, 0.56)—note that the green value exceeds the normalized value of 1. From this illustration it is obvious that there are many combinations of Y, Cb, and Cr that result in invalid RGB values; these YCbCr values must be processed so as to generate valid RGB values. Figure 3.16 shows the RGB normalized limits transformed into the YCbCr color space. Best results are obtained using a constant luma and constant hue approach—Y is not
altered while Cb and Cr are limited to the maximum valid values having the same hue as the invalid color prior to limiting. The constant hue principle corresponds to moving invalid CbCr combinations directly towards the CbCr origin (128, 128), until they lie on the surface of the valid YCbCr color block. When converting to the RGB color space from a non-RGB color space, care must be taken to include saturation logic to ensure overflow and underflow wrap-around conditions do not occur due to the finite precision of digital circuitry. 8-bit RGB values less than 0 must be set to 0, and values greater than 255 must be set to 255.
32
Chapter 3: Color Spaces
Gamma Correction OUT
The transfer function of most CRT displays produces an intensity that is proportional to some power (referred to as gamma) of the signal amplitude. As a result, high-intensity ranges are expanded and low-intensity ranges are compressed (see Figure 3.17). This is an advantage in combatting noise, as the eye is approximately equally sensitive to equally relative intensity changes. By “gamma correcting” the video signals before transmission, the intensity output of the display is roughly linear (the gray line in Figure 3.17), and transmission-induced noise is reduced. To minimize noise in the darker areas of the image, modern video systems limit the gain of the curve in the black region. This technique limits the gain close to black and stretches the remainder of the curve to maintain function and tangent continuity. Although video standards assume a display gamma of about 2.2, a gamma of about 2.5 is more realistic for CRT displays. However, this difference improves the viewing in a dimly lit environment. More accurate viewing in a brightly lit environment may be accomplished by applying another gamma factor of about 1.14 (2.5/2.2). It is also common to tweak the gamma curve in the display to get closer to the “film look.”
Early NTSC Systems Early NTSC systems assumed a simple transform at the display, with a gamma of 2.2. RGB values are normalized to have a range of 0 to 1:
1.0
TRANSMITTED PRE-CORRECTION
0.8
0.6
0.4
DISPLAY CHARACTERISTIC
0.2
0.0 0.0
0.2
0.6
0.8
1.0
IN
Figure 3.17. Effect of Gamma.
To compensate for the nonlinear display, linear RGB data was “gamma-corrected” prior to transmission by the inverse transform. RGB values are normalized to have a range of 0 to 1: R´ = R1/2.2
R = R´2.2
G´ = G1/2.2
G = G´2.2
B´ = B1/2.2
B = B´2.2
0.4
Gamma Correction
Early PAL and SECAM Systems Most early PAL and SECAM systems assumed a simple transform at the display, with a gamma of 2.8. RGB values are normalized to have a range of 0 to 1: R = R´2.8 G = G´2.8 B = B´2.8
To compensate for the nonlinear display, linear RGB data was “gamma-corrected” prior to transmission by the inverse transform. RGB values are normalized to have a range of 0 to 1: R´ = R1/2.8 G´ = G1/2.8 B´ = B1/2.8
33
for (R´, G´, B´) ≥ 0.0812 R = ((R´ + 0.099) / 1.099)1/0.45 G = ((G´ + 0.099) / 1.099)1/0.45 B = ((B´ + 0.099) / 1.099)1/0.45
To compensate for the nonlinear display, linear RGB data is “gamma-corrected” prior to transmission by the inverse transform. RGB values are normalized to have a range of 0 to 1: for R, G, B < 0.018 R´ = 4.5 R G´ = 4.5 G B´ = 4.5 B
for R, G, B ≥ 0.018 R´ = 1.099 R0.45 – 0.099 G´ = 1.099 G0.45 – 0.099 B´ = 1.099 B0.45 – 0.099
Current Systems Current NTSC, 480i, 480p and HDTV video systems assume the following transform at the display, with a gamma of [1/0.45]. RGB values are normalized to have a range of 0 to 1: for (R´, G´, B´) < 0.0812 R = R´ / 4.5 G = G´ / 4.5 B = B´ / 4.5
Although most PAL and SECAM standards specify a gamma of 2.8, a value of [1/0.45] is now commonly used. Thus, these equations are also now used for PAL, SECAM, 576i and 576p video systems.
Non-CRT Displays Since they are not based on CRTs, the LCD, LCOS, DLP and plasma displays have different display transforms. To simplify interfacing to these displays, their electronics are designed to accept standard gamma-corrected video and then compensate for the actual transform of the display panel.
34
Chapter 3: Color Spaces
Constant Luminance Problem Due to the wrong order of the gamma and matrix operations, the U and V (or Cb and Cr) signals also contribute to the luminance (Y) signal. This causes an error in the perceived luminance when the amplitude of U and V is not correct. This may be due to bandwidth-limiting U and V or a non-nominal setting of the U and V gain (color saturation). For low color frequencies, there is no problem. For high color frequencies, U and V disappear and consequently R´, G´, and B´ degrade to be equal to (only) Y.
References 1. Benson, K. Blair, Television Engineering Handbook. McGraw-Hill, Inc., 1986. 2. Clarke, C.K.P., 1986, Colour Encoding and Decoding Techniques for Line-Locked Sampled PAL and NTSC Television Signals, BBC Research Department Report BBC RD1986/2.
3. Devereux, V. G., 1987, Limiting of YUV digital video signals, BBC Research Department Report BBC RD1987 22. 4. EIA Standard EIA–189–A, July 1976, Encoded Color Bar Signal. 5. Faroudja, Yves Charles, NTSC and Beyond. IEEE Transactions on Consumer Electronics, Vol. 34, No. 1, February 1988. 6. ITU-R BT.470–6, 1998, Conventional Television Systems. 7. ITU-R BT.601–5, 1995, Studio Encoding Parameters of Digital Television for Standard 4:3 and Widescreen 16:9 Aspect Ratios. 8. ITU-R BT.709–5, 2002, Parameter Values for the HDTV Standards for Production and International Programme Exchange. 9. Photo CD Information Bulletin, Fully Utilizing Photo CD Images–PhotoYCC Color Encoding and Compression Schemes, May 1994, Eastman Kodak Company.
Chapter 4: Video Signals Overview
Digital Component Video Background
35
Chapter 4
Video Signals Overview Video signals come in a wide variety of options—number of scan lines, interlaced vs. progressive, analog vs. digital, and so on. This chapter provides an overview of the common video signal formats and their timing.
Digital Component Video Background In digital component video, the video signals are in digital form (YCbCr or R´G´B´), being encoded to composite NTSC, PAL or SECAM only when it is necessary for broadcasting or recording purposes. The European Broadcasting Union (EBU) became interested in a standard for digital component video due to the difficulties of exchanging video material between the 576i PAL and SECAM systems. The format held the promise that the digital video signals would be identical whether sourced in a PAL or SECAM country, allowing subsequent encoding to the appropriate composite form for broadcasting. Consultations with the Society of Motion Pic-
ture and Television Engineers (SMPTE) resulted in the development of an approach to support international program exchange, including 480i systems. A series of demonstrations was carried out to determine the quality and suitability for signal processing of various methods. From these investigations, the main parameters of the digital component coding, filtering, and timing were chosen and incorporated into ITU-R BT.601. BT.601 has since served as the starting point for other digital component video standards.
Coding Ranges The selection of the coding ranges balanced the requirements of adequate capacity for signals beyond the normal range and minimizing quantizing distortion. Although the black level of a video signal is reasonably well defined, the white level can be subject to variations due to video signal and equipment tolerances. Noise, gain variations, and transients produced by filtering can produce signal levels outside the nominal ranges.
35
36
Chapter 4: Video Signals Overview
8 or 10 bits per sample are used for each of the YCbCr or R´G´B´ components. Although 8bit coding introduces some quantizing distortion, it was originally felt that most video sources contained sufficient noise to mask most of the quantizing distortion. However, if the video source is virtually noise-free, the quantizing distortion is noticeable as contouring in areas where the signal brightness gradually changes. In addition, at least two additional bits of fractional YCbCr or R´G´B´ data were desirable to reduce rounding effects when transmitting between equipment in the studio editing environment. For these reasons, most pro-video equipment uses 10-bit YCbCr or R´G´B´, allowing 2 bits of fractional YCbCr or R´G´B´ data to be maintained. Initial proposals had equal coding ranges for all three YCbCr components. However, this was changed so that Y had a greater margin for overloads at the white levels, as white level limiting is more visible than black. Thus, the nominal 8-bit Y levels are 16–235, while the nominal 8-bit CbCr levels are 16–240 (with 128 corresponding to no color). Occasional excursions into the other levels are permissible, but never at the 0 and 255 levels. For 8-bit systems, the values of 00H and FFH are reserved for timing information. For 10-bit systems, the values of 000H–003H and 3FCH–3FFH are reserved for timing information, to maintain compatibility with 8-bit systems. The YCbCr or R´G´B´ levels to generate 75% color bars are discussed in Chapter 3. Digital R´G´B´ signals are defined to have the same nominal levels as Y to provide processing margin and simplify the digital matrix conversions between R´G´B´ and YCbCr.
SDTV Sample Rate Selection Line-locked sampling of analog R´G´B´ or YUV video signals is done. This technique produces a static orthogonal sampling grid in which samples on the current scan line fall directly beneath those on previous scan lines and fields, as shown Figures 3.2 through 3.11. Another important feature is that the sampling is locked in phase so that one sample is coincident with the 50% amplitude point of the falling edge of analog horizontal sync (0H). This ensures that different sources produce samples at nominally the same positions in the picture. Making this feature common simplifies conversion from one standard to another. For 480i and 576i video systems, several Y sampling frequencies were initially examined, including four times Fsc. However, the fourtimes Fsc sampling rates did not support the requirement of simplifying international exchange of programs, so they were dropped in favor of a single common sampling rate. Because the lowest sample rate possible (while still supporting quality video) was a goal, a 12MHz sample rate was preferred for a long time, but eventually was considered to be too close to the Nyquist limit, complicating the filtering requirements. When the frequencies between 12 MHz and 14.3 MHz were examined, it became evident that a 13.5-MHz sample rate for Y provided some commonality between 480i and 576i systems. Cb and Cr, being color difference signals, do not require the same bandwidth as the Y, so may be sampled at one-half the Y sample rate, or 6.75 MHz. The “4:2:2” notation now commonly used originally applied to NTSC and PAL video,
480i and 480p Systems
implying that Y, U and V were sampled at 4×, 2× and 2× the color subcarrier frequency, respectively. The “4:2:2” notation was then adapted to BT.601 digital component video, implying that the sampling frequencies of Y, Cb and Cr were 4×, 2× and 2× 3.375 MHz, respectively. “4:2:2” now commonly means that the sample rate of Cb and Cr is one-half that of Y, regardless of the actual sample rates used. With 13.5-MHz sampling, each scan line contains 858 samples (480i systems) or 864 samples (576i systems) and consists of a digital blanking interval followed by an active line period. Both the 480i and 576i systems use 720 samples during the active line period. Having a common number of samples for the active line period simplifies the design of multistandard equipment and standards conversion. With a sample rate of 6.75 MHz for Cb and Cr (4:2:2 sampling), each active line period contains 360 Cr samples and 360 Cb samples. With analog systems, problems may arise with repeated processing, causing an extension of the blanking intervals and softening of the blanking edges. Using 720 digital samples for the active line period accommodates the range of analog blanking tolerances of both the 480i and 576i systems. Therefore, repeated processing may be done without affecting the digital blanking interval. Blanking to define the analog picture width need only be done once, preferably at the display or upon conversion to analog video. Initially, BT.601 supported only 480i and 576i systems with a 4:3 aspect ratio (720 × 480i and 720 × 576i active resolutions). Support for a 16:9 aspect ratio was then added (960 × 480i and 960 × 576i active resolutions) using an 18 MHz sample rate.
37
EDTV Sample Rate Selection ITU BT.1358 defines the progressive SDTV video signals, also known as 480p or 576p, or Enhanced Digital Television (EDTV). The sample rate is doubled to 27 MHz (4:3 aspect ratio) or 36 MHz (16:9 aspect ratio) in order to keep the same static orthogonal sampling grid as that used by BT.601.
HDTV Sample Rate Selection ITU BT.709 defines the 720p, 1080i and 1080p video signals, respectively. With HDTV, a different technique was used—the number of active samples per line and the number of active lines per frame is constant regardless of the frame rate. Thus, in order to keep a static orthogonal sampling grid, each frame rate uses a different sample clock rate.
480i and 480p Systems Interlaced Analog Composite Video (M) NTSC and (M) PAL are analog composite video signals that carry all timing and color information within a single signal. These analog interfaces use 525 lines per frame and are discussed in detail in Chapter 8.
Interlaced Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 480i (since there are typically 480 active scan lines per frame and they are interlaced), the frame rate is usually 29.97 Hz (30/1.001) for compatibility with (M) NTSC timing. The analog interface uses 525 lines per frame, with active video present on lines 23– 262 and 286–525, as shown in Figure 4.1.
38
Chapter 4: Video Signals Overview
START OF VSYNC
523
524
261
525
262
1
263
2
264
3
265
4
266
5
267
6
268
7
269
8
270
9
271
23
272
HSYNC
H/2
10
285
286
HSYNC / 2
H/2
H/2
H/2
Figure 4.1. 480i Vertical Interval Timing.
START OF VSYNC
523
524
525
1
7
8
13
14
15
Figure 4.2. 480p Vertical Interval Timing.
16
45
480i and 480p Systems
For the 29.97 Hz frame rate, each scan line time (H) is about 63.556 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
Progressive Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 480p (since there are typically 480 active scan lines per frame and they are progressive), the frame rate is usually 59.94 Hz (60/1.001) for easier compatibility with (M) NTSC timing. The analog interface uses 525 lines per frame, with active video present on lines 45–524, as shown in Figure 4.2. For the 59.94 Hz frame rate, each scan line time (H) is about 31.776 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
39
Interlaced Digital Component Video BT.601 and SMPTE 267M specify the representation for 480i digital R´G´B´ or YCbCr video signals. Active resolutions defined within BT.601 and SMPTE 267M, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 960 × 480i 720 × 480i
18.0 MHz 13.5 MHz
29.97 Hz 29.97 Hz
Other common active resolutions, their 1× sample rates (Fs), and frame rates, are: 864 × 480i 704 × 480i 640 × 480i 544 × 480i 528 × 480i 480 × 480i 352 × 480i
16.38 MHz 13.50 MHz 12.27 MHz 10.12 MHz 9.900 MHz 9.000 MHz 6.750 MHz
SAMPLE RATE = 13.5 MHZ
16 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
138 SAMPLES (0–137)
720 SAMPLES (138–857)
TOTAL LINE 858 SAMPLES (0–857)
Figure 4.3. 480i Analog - Digital Relationship (4:3 Aspect Ratio, 29.97 Hz Refresh, 13.5 MHz Sample Clock).
29.97 Hz 29.97 Hz 29.97 Hz 29.97 Hz 29.97 Hz 29.97 Hz 29.97 Hz
40
Chapter 4: Video Signals Overview
SAMPLE RATE = 18.0 MHZ
21.5 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
184 SAMPLES (0–183)
960 SAMPLES (184–1143)
TOTAL LINE 1144 SAMPLES (0–1143)
Figure 4.4. 480i Analog - Digital Relationship (16:9 Aspect Ratio, 29.97 Hz Refresh, 18 MHz Sample Clock).
SAMPLE RATE = 12.27 MHZ
14.5 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
140 SAMPLES (0–139)
640 SAMPLES (140–779)
TOTAL LINE 780 SAMPLES (0–779)
Figure 4.5. 480i Analog - Digital Relationship (4:3 Aspect Ratio, 29.97 Hz Refresh, 12.27 MHz Sample Clock).
480i and 480p Systems
SAMPLE RATE = 10.125 MHZ
12 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
99 SAMPLES (0–98)
544 SAMPLES (99–642)
TOTAL LINE 643 SAMPLES (0–642)
Figure 4.6. 480i Analog - Digital Relationship (4:3 Aspect Ratio, 29.97 Hz Refresh, 10.125 MHz Sample Clock).
SAMPLE RATE = 9 MHZ
10.7 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
92 SAMPLES (0–91)
480 SAMPLES (92–571)
TOTAL LINE 572 SAMPLES (0–571)
Figure 4.7. 480i Analog - Digital Relationship (4:3 Aspect Ratio, 29.97 Hz Refresh, 9 MHz Sample Clock).
41
42
Chapter 4: Video Signals Overview
LINE 1 (V = 1) LINE 4
BLANKING LINE 23 (V = 0)
FIELD 1 (F = 0) ODD
F
V
1–3
1
1
4–22
0
1
23–262
0
0
263–265
0
1
266–285
1
1
286–525
1
0
LINE FIELD 1 ACTIVE VIDEO
NUMBER
LINE 263 (V = 1) LINE 266 BLANKING LINE 286 (V = 0)
FIELD 2 (F = 1) EVEN
FIELD 2 ACTIVE VIDEO
LINE 525 (V = 0) LINE 3
H = 1 EAV H = 0 SAV
Figure 4.8. 480i Digital Vertical Timing (480 Active Lines). F and V change state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.3 through 4.7. These active lines are used by the SMPTE RP-202, ATSC A/54a and ARIB STD-B32 standards. EIA–861B (DVI and HDMI timing) specifies lines 22–261 and 285–524 for active video. IEC 61834-2, ITU-R BT.1618 and SMPTE 314M (DV formats) specify lines 23–262 and 285–524 for active video. ITU-R BT.656 specifies lines 20–263 and 283–525 for active video, resulting in 487 total active lines per frame.
480i and 480p Systems
864 × 480i is a 16:9 square pixel format, while 640 × 480i is a 4:3 square pixel format. Although the ideal 16:9 resolution is 854 × 480i, 864 × 480i supports the MPEG 16 × 16 block structure. The 704 × 480i format is done by using the 720 × 480i format, and blanking the first eight and last eight samples each active scan line. Example relationships between the analog and digital signals are shown in Figures 4.3 through 4.7. The H (horizontal blanking), V (vertical blanking), and F (field) signals are defined in Figure 4.8. The H, V and F timing indicated is compatible with video compression standards rather than BT.656 discussed in Chapter 6.
Progressive Digital Component Video BT.1358 and SMPTE 293M specify the representation for 480p digital R´G´B´ or YCbCr video signals. Active resolutions defined within
43
BT.1358 and SMPTE 293M, their 1× sample rates (Fs), and frame rates, are: 960 × 480p 720 × 480p
36.0 MHz 27.0 MHz
59.94 Hz 59.94 Hz
Other common active resolutions, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 864 × 480p 704 × 480p 640 × 480p 544 × 480p 528 × 480p 480 × 480p 352 × 480p
32.75 MHz 27.00 MHz 24.54 MHz 20.25 MHz 19.80 MHz 18.00 MHz 13.50 MHz
59.94 Hz 59.94 Hz 59.94 Hz 59.94 Hz 59.94 Hz 59.94 Hz 59.94 Hz
864 × 480p is a 16:9 square pixel format, while 640 × 480p is a 4:3 square pixel format. Although the ideal 16:9 resolution is 854 × 480p, 864 × 480p supports the MPEG 16 × 16 block structure. The 704 × 480p format is done
SAMPLE RATE = 27.0 MHZ
16 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
138 SAMPLES (0–137)
720 SAMPLES (138–857)
TOTAL LINE 858 SAMPLES (0–857)
Figure 4.9. 480p Analog - Digital Relationship (4:3 Aspect Ratio, 59.94 Hz Refresh, 27 MHz Sample Clock).
44
Chapter 4: Video Signals Overview
SAMPLE RATE = 36.0 MHZ
21.5 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
184 SAMPLES (0–183)
960 SAMPLES (184–1143)
TOTAL LINE 1144 SAMPLES (0–1143)
Figure 4.10. 480p Analog - Digital Relationship (16:9 Aspect Ratio, 59.94 Hz Refresh, 36 MHz Sample Clock).
SAMPLE RATE = 24.54 MHZ
14.5 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
140 SAMPLES (0–139)
640 SAMPLES (140–779)
TOTAL LINE 780 SAMPLES (0–779)
Figure 4.11. 480p Analog - Digital Relationship (4:3 Aspect Ratio, 59.94 Hz Refresh, 24.54 MHz Sample Clock).
480i and 480p Systems
45
SAMPLE RATE = 20.25 MHZ
12 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
99 SAMPLES (0–98)
544 SAMPLES (99–642)
TOTAL LINE 643 SAMPLES (0–642)
Figure 4.12. 480p Analog - Digital Relationship (4:3 Aspect Ratio, 59.94 Hz Refresh, 20.25 MHz Sample Clock).
LINE 1 (V = 1) BLANKING
LINE 45 (V = 0) LINE
F
V
1–44
0
1
45–524 525
0 0
0 1
NUMBER
ACTIVE VIDEO
LINE 525 (V = 1)
H = 1 EAV H = 0 SAV
Figure 4.13. 480p Digital Vertical Timing (480 Active Lines). V changes state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.9 through 4.12. These active lines are used by the SMPTE RP-202, ATSC A/54a and ARIB STD-B32 standards. However, EIA–861B (DVI and HDMI timing) specifies lines 43–522 for active video.
46
Chapter 4: Video Signals Overview
by using the 720 × 480p format, and blanking the first eight and last eight samples each active scan line. Example relationships between the analog and digital signals are shown in Figures 4.9 through 4.12. The H (horizontal blanking), V (vertical blanking), and F (field) signals are defined in Figure 4.13. The H, V and F timing indicated is compatible with video compression standards rather than BT.656 discussed in Chapter 6.
SIF and QSIF SIF is defined to have an active resolution of 352 × 240p. Square pixel SIF is defined to have an active resolution of 320 × 240p. QSIF is defined to have an active resolution of 176 × 120p. Square pixel QSIF is defined to have an active resolution of 160 × 120p.
576i and 576p Systems Interlaced Analog Composite Video (B, D, G, H, I, N, NC) PAL are analog composite video signals that carry all timing and color information within a single signal. These analog interfaces use 625 lines per frame and are discussed in detail in Chapter 8.
Interlaced Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 576i (since there are typically 576 active scan lines per frame and they are interlaced), the frame rate is usually 25 Hz for compatibility with PAL timing. The analog interface uses 625 lines per frame, with active video present on lines 23–310 and 336–623, as shown in Figure 4.14.
For the 25 Hz frame rate, each scan line time (H) is 64 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
Progressive Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 576p (since there are typically 576 active scan lines per frame and they are progressive), the frame rate is usually 50 Hz for compatibility with PAL timing. The analog interface uses 625 lines per frame, with active video present on lines 45–620, as shown in Figure 4.15. For the 50 Hz frame rate, each scan line time (H) is 32 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
Interlaced Digital Component Video BT.601 specifies the representation for 576i digital R´G´B´ or YCbCr video signals. Active resolutions defined within BT.601, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 960 × 576i 720 × 576i
18.0 MHz 13.5 MHz
25 Hz 25 Hz
Other common active resolutions, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1024 × 576i 768 × 576i 704 × 576i 544 × 576i 480 × 576i
19.67 MHz 14.75 MHz 13.50 MHz 10.12 MHz 9.000 MHz
25 Hz 25 Hz 25 Hz 25 Hz 25 Hz
576i and 576p Systems
47
START OF VSYNC
620
621
308
622
309
623
310
624
311
625
312
1
313
2
314
3
315
4
316
5
6
317
318
HSYNC
H/2
7
319
23
320
24
336
HSYNC / 2
H/2
H/2
H/2
Figure 4.14. 576i Vertical Interval Timing.
START OF VSYNC
619
620
621
625
1
2
6
7
8
Figure 4.15. 576p Vertical Interval Timing.
9
45
337
48
Chapter 4: Video Signals Overview
SAMPLE RATE = 13.5 MHZ
12 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
144 SAMPLES (0–143)
720 SAMPLES (144–863)
TOTAL LINE 864 SAMPLES (0–863)
Figure 4.16. 576i Analog - Digital Relationship (4:3 Aspect Ratio, 25 Hz Refresh, 13.5 MHz Sample Clock).
SAMPLE RATE = 18.0 MHZ
16 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
192 SAMPLES (0–191)
960 SAMPLES (192–1151)
TOTAL LINE 1152 SAMPLES (0–1151)
Figure 4.17. 576i Analog - Digital Relationship (16:9 Aspect Ratio, 25 Hz Refresh, 18 MHz Sample Clock).
576i and 576p Systems
SAMPLE RATE = 14.75 MHZ
13 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
176 SAMPLES (0–175)
768 SAMPLES (176–943)
TOTAL LINE 944 SAMPLES (0–943)
Figure 4.18. 576i Analog - Digital Relationship (4:3 Aspect Ratio, 25 Hz Refresh, 14.75 MHz Sample Clock).
SAMPLE RATE = 10.125 MHZ
9 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
104 SAMPLES (0–103)
544 SAMPLES (104–647)
TOTAL LINE 648 SAMPLES (0–647)
Figure 4.19. 576i Analog - Digital Relationship (4:3 Aspect Ratio, 25 Hz Refresh, 10.125 MHz Sample Clock).
49
50
Chapter 4: Video Signals Overview
LINE 1
LINE 1 (V = 1) BLANKING LINE 23 (V = 0)
FIELD 1 (F = 0) EVEN
F
V
1–22
0
1
23–310
0
0
311–312
0
1
313–335
1
1
336–623
1
0
624–625
1
1
LINE FIELD 1 ACTIVE VIDEO
NUMBER
LINE 311 (V = 1) LINE 313 BLANKING LINE 336 (V = 0)
FIELD 2 (F = 1) ODD
FIELD 2 ACTIVE VIDEO
LINE 624 (V = 1) BLANKING LINE 625 (V = 1)
LINE 625
H = 1 EAV H = 0 SAV
Figure 4.20. 576i Digital Vertical Timing (576 Active Lines). F and V change state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.16 through 4.19. IEC 61834-2, ITU-R BT.1618 and SMPTE 314M (DV formats) specify lines 23–310 and 335–622 for active video.
576i and 576p Systems
1024 × 576i is a 16:9 square pixel format, while 768 × 576i is a 4:3 square pixel format. The 704 × 576i format is done by using the 720 × 576i format, and blanking the first eight and last eight samples each active scan line. Example relationships between the analog and digital signals are shown in Figures 4.16 through 4.19. The H (horizontal blanking), V (vertical blanking), and F (field) signals are defined in Figure 4.20. The H, V and F timing indicated is compatible with video compression standards rather than BT.656 discussed in Chapter 6.
960 × 576p 720 × 576p
36.0 MHz 27.0 MHz
51
50 Hz 50 Hz
Other common active resolutions, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1024 × 576p 768 × 576p 704 × 576p 544 × 576p 480 × 576p
39.33 MHz 29.50 MHz 27.00 MHz 20.25 MHz 18.00 MHz
50 Hz 50 Hz 50 Hz 50 Hz 50 Hz
1024 × 576p is a 16:9 square pixel format, while 768 × 576p is a 4:3 square pixel format. The 704 × 576p format is done by using the 720 × 576p format, and blanking the first eight and last eight samples each active scan line. Example relationships between the analog and digital signals are shown in Figures 4.21 through 4.24.
Progressive Digital Component Video BT.1358 specifies the representation for 576p digital R´G´B´ or YCbCr signals. Active resolutions defined within BT.1358, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are:
SAMPLE RATE = 27.0 MHZ
12 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
144 SAMPLES (0–143)
720 SAMPLES (144–863)
TOTAL LINE 864 SAMPLES (0–863)
Figure 4.21. 576p Analog - Digital Relationship (4:3 Aspect Ratio, 50 Hz Refresh, 27 MHz Sample Clock).
52
Chapter 4: Video Signals Overview
SAMPLE RATE = 36.0 MHZ
16 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
192 SAMPLES (0–191)
960 SAMPLES (192–1151)
TOTAL LINE 1152 SAMPLES (0–1151)
Figure 4.22. 576p Analog - Digital Relationship (16:9 Aspect Ratio, 50 Hz Refresh, 36 MHz Sample Clock).
SAMPLE RATE = 29.5 MHZ
13 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
176 SAMPLES (0–175)
768 SAMPLES (176–943)
TOTAL LINE 944 SAMPLES (0–943)
Figure 4.23. 576p Analog - Digital Relationship (4:3 Aspect Ratio, 50 Hz Refresh, 29.5 MHz Sample Clock).
576i and 576p Systems
53
SAMPLE RATE = 20.25 MHZ
9 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
104 SAMPLES (0–103)
544 SAMPLES (104–647)
TOTAL LINE 648 SAMPLES (0–647)
Figure 4.24. 576p Analog - Digital Relationship (4:3 Aspect Ratio, 50 Hz Refresh, 20.25 MHz Sample Clock).
LINE 1 (V = 1) BLANKING
LINE 45 (V = 0) F
V
1–44
0
1
45–620
0
0
621–625
0
1
LINE NUMBER ACTIVE VIDEO
LINE 621 (V = 1) BLANKING
LINE 625 (V = 1)
H = 1 EAV H = 0 SAV
Figure 4.25. 576p Digital Vertical Timing (576 Active Lines). V changes state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.21 through 4.24.
54
Chapter 4: Video Signals Overview
The H (horizontal blanking), V (vertical blanking), and F (field) signals are defined in Figure 4.25. The H, V and F timing indicated is compatible with video compression standards rather than BT.656 discussed in Chapter 6.
Progressive Digital Component Video SMPTE 296M specifies the representation for 720p digital R´G´B´ or YCbCr signals. Active resolutions defined within SMPTE 296M, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p
720p Systems Progressive Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 720p (since there are typically 720 active scan lines per frame and they are progressive), the frame rate is usually 59.94 Hz (60/1.001) to simplify the generation of (M) NTSC video. The analog interface uses 750 lines per frame, with active video present on lines 26–745, as shown in Figure 4.26. For the 59.94 Hz frame rate, each scan line time (H) is about 22.24 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
74.176 MHz 74.250 MHz 74.250 MHz 74.176 MHz 74.250 MHz 74.250 MHz 74.176 MHz 74.250 MHz
23.976 Hz 24.000 Hz 25.000 Hz 29.970 Hz 30.000 Hz 50.000 Hz 59.940 Hz 60.000 Hz
Note that square pixels and a 16:9 aspect ratio are used. Example relationships between the analog and digital signals are shown in Figures 4.27 and 4.28, and Table 4.1. The H (horizontal blanking), V (vertical blanking), and F (field) signals are as defined in Figure 4.29.
START OF VSYNC
744
745
746
750
1
2
6
7
8
Figure 4.26. 720p Vertical Interval Timing.
9
26
720p Systems
SAMPLE RATE = 74.176 OR 74.25 MHZ
110 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
370 SAMPLES (0–369)
1280 SAMPLES (370–1649) TOTAL LINE 1650 SAMPLES (0–1649)
Figure 4.27. 720p Analog - Digital Relationship (16:9 Aspect Ratio, 59.94 Hz Refresh, 74.176 MHz Sample Clock and 60 Hz Refresh, 74.25 MHz Sample Clock).
[C] SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
[B] SAMPLES
1280 SAMPLES TOTAL LINE [A] SAMPLES
Figure 4.28. General 720p Analog - Digital Relationship.
55
56
Chapter 4: Video Signals Overview
Active Horizontal Samples
1280
1× Y Sample Rate (MHz)
Frame Rate (Hz)
Total Horizontal Samples (A)
Horizontal Blanking Samples (B)
C Samples
24/1.001
74.25/1.001
4125
2845
2585
24
74.25
4125
2845
2585
251
48
1536
256
21
251
49.5
1584
304
25
25
74.25
3960
2680
2420
30/1.001
74.25/1.001
3300
2020
1760
30
74.25
3300
2020
1760
50
74.25
1980
700
440
60/1.001
74.25/1.001
1650
370
110
60
74.25
1650
370
110
Notes: 1. Useful for CRT-based 50 Hz HDTVs based on a 31.250 kHz horizontal frequency. Sync pulses are –300 mV bi-level, rather than ±300 mV tri-level. 720p content scaled vertically to 1152i active scan lines; 1250i total scan lines instead of 750p.
Table 4.1. Various 720p Analog - Digital Parameters for Figure 4.28. LINE 1 (V = 1) BLANKING
LINE 26 (V = 0) F
V
1–25
0
1
26–745
0
0
746–750
0
1
LINE NUMBER ACTIVE VIDEO
LINE 746 (V = 1) BLANKING
LINE 750 (V = 1)
H = 1 EAV H = 0 SAV
Figure 4.29. 720p Digital Vertical Timing (720 Active Lines). V changes state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.27 and 4.28.
1080i and 1080p Systems
1080i and 1080p Systems Interlaced Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 1080i (since there are typically 1080 active scan lines per frame and they are interlaced), the frame rate is usually 25 or 29.97 Hz (30/1.001) to simplify the generation of (B, D, G, H, I) PAL or (M) NTSC video. The analog interface uses 1125 lines per frame, with active video present on lines 21–560 and 584–1123, as shown in Figure 4.30. MPEG-2 and MPEG-4 systems use 1088 lines, rather than 1080, in order to have a multiple of 32 scan lines per frame. In this case, an additional 4 lines per field after the active video are used. For the 25 Hz frame rate, each scan line time is about 35.56 µs. For the 29.97 Hz frame rate, each scan line time is about 29.66 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5. 1152i Format The 1152i active (1250 total) line format is not a broadcast transmission format. However, it is being used as an analog interconnection standard from HD set-top boxes and DVD players to 50 Hz CRT-based HDTVs. This enables 50 Hz HDTVs to use a fixed 31.25 kHz horizontal frequency, reducing their cost. Other HDTV display technologies, such as DLP, LCD and plasma, are capable of handling the native timing of 720p50 (750p50 with VBI) and 1080i25 (1125i25 with VBI) analog signals.
57
The set-top box or DVD player converts 720p50 and 1080i25 content to the 1152i25 format. 1280 × 720p50 content is scaled to 1280 × 1152i25; 1920 × 1080i25 content is presented letterboxed in a 1920 × 1152i25 format. HDTVs will have a nominal vertical zoom mode for correcting the geometry of 1080i25, which can be recognized by the vertical synchronizing signal.
Progressive Analog Component Video Analog component signals are comprised of three signals, analog R´G´B´ or YPbPr. Referred to as 1080p (since there are typically 1080 active scan lines per frame and they are progressive), the frame rate is usually 50 or 59.94 Hz (60/1.001) to simplify the generation of (B, D, G, H, I) PAL or (M) NTSC video. The analog interface uses 1125 lines per frame, with active video present on lines 42–1121, as shown in Figure 4.31. MPEG-2 and MPEG-4 systems use 1088 lines, rather than 1080, in order to have a multiple of 16 scan lines per frame. In this case, an additional 8 lines per frame after the active video are used. For the 50 Hz frame rate, each scan line time is about 17.78 µs. For the 59.94 Hz frame rate, each scan line time is about 14.83 µs. Detailed horizontal timing is dependent on the specific video interface used, as discussed in Chapter 5.
58
Chapter 4: Video Signals Overview
1123
1125
560
562
1
2
563
564
3
565
4
566
5
6
567
568
7
569
21
584
START OF VSYNC
Figure 4.30. 1080i Vertical Interval Timing.
START OF VSYNC
1120
1121
1122
1125
1
2
6
7
8
Figure 4.31. 1080p Vertical Interval Timing.
9
42
1080i and 1080p Systems
SAMPLE RATE = 74.25 OR 74.176 MHZ
88 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
280 SAMPLES (0–279)
1920 SAMPLES (280–2199) TOTAL LINE 2200 SAMPLES (0–2199)
Figure 4.32. 1080i Analog - Digital Relationship (16:9 Aspect Ratio, 29.97 Hz Refresh, 74.176 MHz Sample Clock and 30 Hz Refresh, 74.25 MHz Sample Clock).
[D] SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
[C] SAMPLES
[A] SAMPLES TOTAL LINE [B] SAMPLES
Figure 4.33. General 1080i Analog - Digital Relationship.
59
60
Chapter 4: Video Signals Overview
Active Horizontal Samples (A)
1920
1440
1280
1× Y Sample Rate (MHz)
Frame Rate (Hz)
Total Horizontal Samples (B)
Horizontal Blanking Samples (C)
D Samples
251
72
2304
384
32
251
74.25
2376
456
38
25
74.25
2640
720
528
30/1.001
74.25/1.001
2200
280
88
30
74.25
2200
280
88
251
54
1728
288
24
25
55.6875
1980
540
396
30/1.001
55.6875/1.001
1650
210
66
30
55.6875
1650
210
66
251
48
1536
256
21
25
49.5
1760
480
352
30/1.001
49.5/1.001
1466.7
186.7
58.7
30
49.5
1466.7
186.7
58.7
Notes: 1. Useful for CRT-based 50 Hz HDTVs based on a 31.250 kHz horizontal frequency. Sync pulses are –300 mV bi-level, rather than ±300 mV tri-level. 1080i content letterboxed in 1152i active scan lines; 1250i total scan lines instead of 1125i.
Table 4.2. Various 1080i Analog - Digital Parameters for Figure 4.33.
Interlaced Digital Component Video ITU-R BT.709 and SMPTE 274M specify the digital component format for the 1080i digital R´G´B´ or YCbCr signal. Active resolutions defined within BT.709 and SMPTE 274M, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1920 × 1080i 1920 × 1080i 1920 × 1080i
74.250 MHz 74.176 MHz 74.250 MHz
25.00 Hz 29.97 Hz 30.00 Hz
Note that square pixels and a 16:9 aspect ratio are used. Other common active resolutions, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are:
1280 × 1080i 1280 × 1080i 1280 × 1080i 1440 × 1080i 1440 × 1080i 1440 × 1080i
49.500 MHz 49.451 MHz 49.500 MHz 55.688 MHz 55.632 MHz 55.688 MHz
25.00 Hz 29.97 Hz 30.00 Hz 25.00 Hz 29.97 Hz 30.00 Hz
Example relationships between the analog and digital signals are shown in Figures 4.32 and 4.33, and Table 4.2. The H (horizontal blanking) and V (vertical blanking) signals are as defined in Figure 4.34.
1080i and 1080p Systems
LINE 1
61
LINE 1 (V = 1) BLANKING
LINE 21 (V = 0) FIELD 1 (F = 0) ODD
F
V
1–20
0
1
21–560
0
0
561–562
0
1
563–583
1
1
584–1123
1
0
1124–1125
1
1
LINE
FIELD 1 ACTIVE VIDEO
NUMBER
LINE 561 (V = 1) LINE 583 BLANKING
LINE 584 (V = 0)
FIELD 2 (F = 1) EVEN
FIELD 2 ACTIVE VIDEO
LINE 1124 (V = 1) BLANKING
LINE 1125
LINE 1125 (V = 1)
H = 1 EAV H = 0 SAV
Figure 4.34. 1080i Digital Vertical Timing (1080 Active Lines). F and V change state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.32 and 4.33.
62
Chapter 4: Video Signals Overview
Progressive Digital Component Video ITU-R BT.709 and SMPTE 274M specify the digital component format for the 1080p digital R´G´B´ or YCbCr signal. Active resolutions defined within BT.709 and SMPTE 274M, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p
74.176 MHz 74.250 MHz 74.250 MHz 74.176 MHz 74.250 MHz 148.50 MHz 148.35 MHz 148.50 MHz
23.976 Hz 24.000 Hz 25.000 Hz 29.970 Hz 30.000 Hz 50.000 Hz 59.940 Hz 60.000 Hz
Note that square pixels and a 16:9 aspect ratio are used. Other common active resolutions, their 1× Y and R´G´B´ sample rates (Fs), and frame rates, are: 1280 × 1080p 1280 × 1080p 1280 × 1080p 1280 × 1080p 1280 × 1080p 1280 × 1080p 1280 × 1080p 1280 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p 1440 × 1080p
49.451 MHz 49.500 MHz 49.500 MHz 49.451 MHz 49.500 MHz 99.000 MHz 98.901 MHz 99.000 MHz 55.632 MHz 55.688 MHz 55.688 MHz 55.632 MHz 55.688 MHz 111.38 MHz 111.26 MHz 111.38 MHz
23.976 Hz 24.000 Hz 25.000 Hz 29.970 Hz 30.000 Hz 50.000 Hz 59.940 Hz 60.000 Hz 23.976 Hz 24.000 Hz 25.000 Hz 29.970 Hz 30.000 Hz 50.000 Hz 59.940 Hz 60.000 Hz
Example relationships between the analog and digital signals are shown in Figures 4.35 and 4.36, and Table 4.3. The H (horizontal blanking), V (vertical blanking), and F (field) signals are as defined in Figure 4.37.
Other Video Systems Some consumer displays, such as those based on LCD and plasma technologies, have adapted other resolutions as their native resolution. Common active resolutions and their names are: 640 × 400 640 × 480 854 × 480 800 × 600 1024 × 768 1280 × 768 1366 × 768 1024 × 1024 1280 × 1024 1600 × 1024 1600 × 1200 1920 × 1200
VGA VGA WVGA SVGA XGA WXGA WXGA XGA SXGA WSXGA UXGA WUXGA
These resolutions, and their timings, are defined for computer monitors by the Video Electronics Standards Association (VESA). Displays based on one of these native resolutions are usually capable of accepting many input resolutions, scaling the source to match the display resolution.
Other Video Systems
SAMPLE RATE = 148.5 OR 148.35 MHZ
88 SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
280 SAMPLES (0–279)
1920 SAMPLES (280–2199) TOTAL LINE 2200 SAMPLES (0–2199)
Figure 4.35. 1080p Analog - Digital Relationship (16:9 Aspect Ratio, 59.94 Hz Refresh, 148.35 MHz Sample Clock and 60 Hz Refresh, 148.5 MHz Sample Clock).
[D] SAMPLES
DIGITAL BLANKING
DIGITAL ACTIVE LINE
[C] SAMPLES
[A] SAMPLES TOTAL LINE [B] SAMPLES
Figure 4.36. General 1080p Analog - Digital Relationship.
63
64
Chapter 4: Video Signals Overview
Active Horizontal Samples (A)
1920
1440
1280
1× Y Sample Rate (MHz)
Frame Rate (Hz)
Total Horizontal Samples (B)
Horizontal Blanking Samples (C)
D Samples
24/1.001
74.25/1.001
2750
830
638
24
74.25
2750
830
638
25
74.25
2640
720
528
30/1.001
74.25/1.001
2200
280
88
30
74.25
2200
280
88
50
148.5
2640
720
528
60/1.001
148.5/1.001
2200
280
88
60
148.5
2200
280
88
24/1.001
55.6875/1.001
2062.5
622.5
478.5
24
55.6875
2062.5
622.5
478.5
25
55.6875
1980
540
396
30/1.001
55.6875/1.001
1650
210
66
30
55.6875
1650
210
66
50
111.375
1980
540
396
60/1.001
111.375/1.001
1650
210
66
60
111.375
1650
210
66
24/1.001
49.5/1.001
1833.3
553.3
425.3
24
49.5
1833.3
553.3
425.3
25
49.5
1760
480
352
30/1.001
49.5/1.001
1466.7
186.7
58.7
30
49.5
1466.7
186.7
58.7
50
99
1760
480
352
60/1.001
99/1.001
1466.7
186.7
58.7
60
99
1466.7
186.7
58.7
Table 4.3. Various 1080p Analog - Digital Parameters for Figure 4.36.
References
65
LINE 1 (V = 1) BLANKING
LINE 42 (V = 0) F
V
1–41
0
1
42–1121
0
0
1122–1125
0
1
LINE NUMBER ACTIVE VIDEO
LINE 1122 (V = 1) BLANKING
LINE 1125 (V = 1)
H = 1 EAV H = 0 SAV
Figure 4.37. 1080p Digital Vertical Timing (1080 Active Lines). V changes state at the EAV sequence at the beginning of the digital line. Note that the digital line number changes state prior to start of horizontal sync, as shown in Figures 4.35 and 4.36.
References 1. EIA–770.1, Analog 525-Line Component Video Interface—Three Channels, November 2001. 2. EIA–770.2, Standard Definition TV Analog Component Video Interface, November 2001. 3. EIA–770.3, High Definition TV Analog Component Video Interface, November 2001. 4. EIA/CEA–861B, A DTV Profile for Uncompressed High Speed Digital Interfaces, May 2002. 5. ITU-R BT.601–5, 1995, Studio Encoding Parameters of Digital Television for Standard 4:3 and Widescreen 16:9 Aspect Ratios. 6. ITU-R BT.709–5, 2002, Parameter Values for the HDTV Standards for Production and International Programme Exchange.
7. ITU-R BT.1358, 1998, Studio Parameters of 625 and 525 Line Progressive Scan Television Systems. 8. SMPTE 267M–1995, Television—Bit-Parallel Digital Interface—Component Video Signal 4:2:2 16 × 9 Aspect Ratio. 9. SMPTE 274M–2003, Television—1920 × 1080 Image Sample Structure, Digital Representation and Digital Timing Reference Sequences for Multiple Picture Rates. 10. SMPTE 293M–2003, Television—720 × 483 Active Line at 59.94-Hz Progressive Scan Production—Digital Representation. 11. SMPTE 296M–2001, Television—1280 × 720 Progressive Image Sample Structure, Analog and Digital Representation and Analog Interface.
66
Chapter 5: Analog Video Interfaces
Chapter 5: Analog Video Interfaces Chapter 5
Analog Video Interfaces For years, the primary video signal used by the consumer market has been composite NTSC or PAL video (Figures 8.2 and 8.13). Attempts have been made to support S-Video, but, until recently, it has been largely limited to S-VHS VCRs and high-end televisions. With the introduction of DVD players, digital set-top boxes, and DTV, there has been renewed interest in providing high-quality video to the consumer market. This equipment not only supports very high quality composite and S-Video signals, but many devices also allow the option of using analog R´G´B´ or YPbPr video. Using analog R´G´B´ or YPbPr video eliminates NTSC/PAL encoding and decoding artifacts. As a result, the picture is sharper and has less noise. More color bandwidth is also available, increasing the horizontal detail.
66
S-Video Interface The RCA phono connector (consumer market) or BNC connector (pro-video market) transfers a composite NTSC or PAL video signal, made by adding the intensity (Y) and color (C) video signals together. The television then has to separate these Y and C video signals in order to display the picture. The problem is that the Y/C separation process is never perfect, as discussed in Chapter 9. Many video components now support a 4pin “S1” S-Video connector, illustrated in Figure 5.1 (the female connector viewpoint). This connector keeps the intensity (Y) and color (C) video signals separate, eliminating the Y/C separation process in the TV. As a result, the picture is sharper and has less noise. Figures 9.2 and 9.3 illustrate the Y signal, and Figures 9.10 and 9.11 illustrate the C signal. NTSC and PAL VBI (vertical blanking interval) data, discussed in Chapter 8, may be present on the 480i or 576i Y video signal.
67
SCART Interface
The “S2” version adds a +5V DC offset to the C signal when a widescreen (16:9) anamorphic program (horizontally squeezed by 25%) is present. A 16:9 TV detects the DC offset and horizontally expands the 4:3 image to fill the screen, restoring the correct aspect ratio of the program. The “S3” version also supports using a +2.3V offset when a program is letterboxed. The IEC 60933-5 standard specifies the SVideo connector, including signal levels.
SCART Interface Most consumer video components in Europe support one or two 21-pin SCART connectors (also known as Peritel, Peritelevision, and Euroconnector). This connection allows analog R´G´B´ video or S-Video, composite video, and analog stereo audio to be transmitted between equipment using a single cable. The composite video signal must always be present, as it provides the basic video timing for the analog R´G´B´ video signals. Note that the 700 mV R´G´B´ signals do not have a blanking pedestal or sync information, as illustrated in Figure 5.4. PAL VBI (vertical blanking interval) data, discussed in Chapter 8, may be present on the 576i composite video signal. There are now several types of SCART pinouts, depending on the specific functions implemented, as shown in Tables 5.1 through 5.3. Pinout details are shown in Figure 5.2. The CENELEC EN 50049–1 and IEC 60933 standards specify the basic SCART connector, including signal levels.
Extended S-Video Interface The PC market also uses an extended SVideo interface. This interface has 7 pins, as shown in Figure 5.1, and is backwards compatible with the 4-pin interface. The use of the three additional pins varies by manufacturer. They may be used to support an I2C interface (SDA bi-directional data pin and SCL clock pin), +12V power, a composite NTSC/PAL video signal (CVBS), or analog R´G´B´ or YPbPr video signals.
7-PIN MINI DIN CONNECTOR
7
4
2
6
4-PIN MINI DIN CONNECTOR
4
3
5
1
2
3
1
1
1, 2 = GND 3 = Y 4 = C 5 = CVBS / SCL (SERIAL CLOCK) 6 = GND / SDA (SERIAL DATA) 7 = NC / +12V
3
5
7
9
11
13
15
17
19
21
1, 2 = GND 3 = Y 4 = C
FIgure 5.1. S-Video Connector and Signal Names.
2
4
6
8
10
12
14
16
18
Figure 5.2. SCART Connector.
20
68
Chapter 5: Analog Video Interfaces
Pin
Function
1 2 3 4 5 6 7 8
right audio out right audio in left / mono audio out ground - for pins 1, 2, 3, 6 ground - for pin 7 left / mono audio in blue (or C) video in / out status and aspect ratio in / out
9 10 11 12 13 14 15 16
ground - for pin 11 data 2 green video in / out data 1 ground - for pin 15 ground - for pin 16 red (or C) video in / out RGB control in / out
17 18 19 20 21
ground - for pin 19 ground - for pin 20 composite (or Y) video out composite (or Y) video in ground - for pins 8, 10, 12, shield
Signal Level
Impedance
0.5V rms 0.5V rms 0.5V rms
< 1K ohm > 10 K ohm < 1K ohm
0.5V rms 0.7V (or 0.3V burst) 9.5V–12V = 4:3 source 4.5V–7V = 16:9 source 0V–2V = inactive source
> 10K ohm 75 ohms > 10K ohm
0.7V
75 ohms
0.7V (or 0.3V burst) 1–3V = RGB, 0–0.4V = composite
75 ohms 75 ohms
1V 1V
75 ohms 75 ohms
Note: Often, the SCART 1 connector supports composite video and RGB, the SCART 2 connector supports composite video and S-Video, and the SCART 3 connector supports only composite video. SCART connections may also be used to add external decoders or descramblers to the video path -- the video signal goes out and comes back in. The RGB control signal controls the TV switch between the composite and RGB inputs, enabling the overlaying of text onto the video, even the internal TV program. This enables an external teletext or closed captioning decoder to add information over the current program. If pin 16 is held high, signifying RGB signals are present, the sync is still carried on the Composite Video pin. Some devices (such DVD players) may provide RGB on a SCART and hold pin 16 permanently high. When a source becomes active, it sets a 12V level on pin 8. This causes the TV to automatically switch to that SCART input. When the source stops, the signal returns to 0V and TV viewing is resumed. If an anamorphic 16:9 program is present, the source raises the signal on pin 8 to only 6V. This causes the TV to switch to that SCART input and at the same time enable the video processing for anamorphic 16:9 programs.
Table 5.1. SCART Connector Signals.
SDTV RGB Interface
SDTV RGB Interface Some SDTV consumer video equipment supports an analog R´G´B´ video interface. NTSC and PAL VBI (vertical blanking interval) data, discussed in Chapter 8, may be present on 480i or 576i R´G´B´ video signals. Three separate RCA phono connectors (consumer market) or BNC connectors (pro-video and PC market) are used. The horizontal and vertical video timing are dependent on the video standard, as discussed in Chapter 4. For sources, the video signal at the connector should have a source impedance of 75Ω ±5%. For receivers, video inputs should be AC-coupled and have a 75-Ω ±5% input impedance. The three signals must be coincident with respect to each other within ±5 ns. Sync information may be present on just the green channel, all three channels, as a separate composite sync signal, or as separate horizontal and vertical sync signals. A gamma of 1/0.45 is used.
7.5 IRE Blanking Pedestal As shown in Figure 5.3, the nominal active video amplitude is 714 mV, including a 7.5 ±2 IRE blanking pedestal. A 286 ±6 mV composite sync signal may be present on just the green channel (consumer market), or all three channels (pro-video market). DC offsets up to ±1V may be present. Analog R´G´B´ Generation Assuming 10-bit D/A converters (DACs) with an output range of 0–1.305V (to match the video DACs used by the NTSC/PAL encoder in Chapter 9), the 10-bit YCbCr to R´G´B´ equations are:
69
R´ = 0.591(Y601 – 64) + 0.810(Cr – 512) G´ = 0.591(Y601 – 64) – 0.413(Cr – 512) – 0.199(Cb – 512) B´ = 0.591(Y601 – 64) + 1.025(Cb – 512)
R´G´B´ has a nominal 10-bit range of 0–518 to match the active video levels used by the NTSC/PAL encoder in Chapter 9. Note that negative values of R´G´B´ should be supported at this point. To implement the 7.5 IRE blanking pedestal, a value of 42 is added to the digital R´G´B´ data during active video. 0 is added during the blanking time. After the blanking pedestal is added, the R´G´B´ data is clamped by a blanking signal that has a raised cosine distribution to slow the slew rate of the start and end of the video signal. For 480i and 576i systems, blank rise and fall times are 140 ±20 ns. For 480p and 576p systems, blank rise and fall times are 70 ±10 ns. Composite sync information may be added to the R´G´B´ data after the blank processing has been performed. Values of 16 (sync present) or 240 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution (between 16 and 240) to slow the slew rate of the sync signal. For 480i and 576i systems, sync rise and fall times are 140 ±20 ns, and horizontal sync width at the 50%-point is 4.7 ±0.1 µs. For 480p and 576p systems, sync rise and fall times are 70 ±10 ns, and horizontal sync width at the 50%point is 2.33 ±0.05 µs. At this point, we have digital R´G´B´ with sync and blanking information, as shown in Figure 5.3 and Table 5.2. The numbers in parentheses in Figure 5.3 indicate the data value for a 10-bit DAC with a full-scale output value of 1.305V. The digital R´G´B´ data drive
70
Chapter 5: Analog Video Interfaces
1.020 V
WHITE LEVEL (800)
100 IRE
0.357 V
7.5 IRE
0.306 V
BLACK LEVEL (282) BLANK LEVEL (240)
40 IRE
0.020 V
SYNC LEVEL (16)
GREEN, BLUE, OR RED CHANNEL, SYNC PRESENT
1.020 V
WHITE LEVEL (800)
100 IRE
0.357 V 0.306 V
7.5 IRE
BLACK LEVEL (282) BLANK LEVEL (240)
GREEN, BLUE, OR RED CHANNEL, NO SYNC PRESENT
FIgure 5.3. SDTV Analog RGB Levels. 7.5 IRE blanking level.
SDTV RGB Interface
WHITE LEVEL (800)
1.020 V
100 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
GREEN, BLUE, OR RED CHANNEL, SYNC PRESENT
WHITE LEVEL (800)
1.020 V
100 IRE
0.321 V
BLACK / BLANK LEVEL (252)
GREEN, BLUE, OR RED CHANNEL, NO SYNC PRESENT
FIgure 5.4. SDTV Analog RGB Levels. 0 IRE blanking level.
71
72
Chapter 5: Analog Video Interfaces
three 10-bit DACs to generate the analog R´G´B´ video signals. As the sample-and-hold action of the DAC introduces a (sin x)/x characteristic, the video data may be digitally filtered by a [(sin x)/x]–1 filter to compensate. Alternately, as an analog lowpass filter is usually present after each DAC, the correction may take place in the analog filter. Video Level
7.5 IRE 0 IRE Blanking Pedestal Blanking Pedestal
white
800
800
black
282
252
blank
240
252
sync
16
16
Table 5.2. SDTV 10-Bit R´G´B´ Values.
0 IRE Blanking Pedestal As shown in Figure 5.4, the nominal active video amplitude is 700 mV, with no blanking pedestal. A 300 ±6 mV composite sync signal may be present on just the green channel (consumer market), or all three channels (provideo market). DC offsets up to ±1V may be present. Analog R´G´B´ Generation Assuming 10-bit DACs with an output range of 0–1.305V (to match the video DACs used by the NTSC/PAL encoder in Chapter 9), the 10-bit YCbCr to R´G´B´ equations are: R´ = 0.625(Y601 – 64) + 0.857(Cr – 512) G´ = 0.625(Y601 – 64) – 0.437(Cr – 512) – 0.210(Cb – 512) B´ = 0.625(Y601 – 64) + 1.084(Cb – 512)
Analog R´G´B´ Digitization Assuming 10-bit A/D converters (ADCs) with an input range of 0–1.305V (to match the video ADCs used by the NTSC/PAL decoder in Chapter 9), the 10-bit R´G´B´ to YCbCr equations are: Y601 = 0.506(R´ – 282) + 0.992(G´ – 282) + 0.193(B´ – 282) + 64 Cb = –0.291(R´ – 282) – 0.573(G´ – 282) + 0.864(B´ – 282) + 512 Cr = 0.864(R´ – 282) – 0.724(G´ – 282) – 0.140(B´ – 282) + 512
R´G´B´ has a nominal 10-bit range of 282– 800 to match the active video levels used by the NTSC/PAL decoder in Chapter 9. Table 5.2 and Figure 5.3 illustrate the 10-bit R´G´B´ values for the white, black, blank, and (optional) sync levels.
R´G´B´ has a nominal 10-bit range of 0–548 to match the active video levels used by the NTSC/PAL encoder in Chapter 9. Note that negative values of R´G´B´ should be supported at this point. The R´G´B´ data is processed as discussed when using a 7.5 IRE blanking pedestal. However, no blanking pedestal is added during active video, and the sync values are 16–252 instead of 16–240. At this point, we have digital R´G´B´ with sync and blanking information, as shown in Figure 5.4 and Table 5.2. The numbers in parentheses in Figure 5.4 indicate the data value for a 10-bit DAC with a full-scale output value of 1.305V. The digital R´G´B´ data drive three 10-bit DACs to generate the analog R´G´B´ video signals.
HDTV RGB Interface
Analog R´G´B´ Digitization Assuming 10-bit ADCs with an input range of 0–1.305V (to match the video ADCs used by the NTSC/PAL decoder in Chapter 9), the 10bit R´G´B´ to YCbCr equations are: Y601 = 0.478(R´ – 252) + 0.938(G´ – 252) + 0.182(B´ – 252) + 64 Cb = –0.275(R´ – 252) – 0.542(G´ – 252) + 0.817(B´ – 252) + 512 Cr = 0.817(R´ – 252) – 0.685(G´ – 252) – 0.132(B´ – 252) + 512
R´G´B´ has a nominal 10-bit range of 252– 800 to match the active video levels used by the NTSC/PAL decoder in Chapter 9. Table 5.2 and Figure 5.4 illustrate the 10-bit R´G´B´ values for the white, black, blank, and (optional) sync levels.
HDTV RGB Interface Some HDTV consumer video equipment supports an analog R´G´B´ video interface. Three separate RCA phono connectors (consumer market) or BNC connectors (pro-video and PC market) are used. The horizontal and vertical video timing are dependent on the video standard, as discussed in Chapter 4. For sources, the video signal at the connector should have a source impedance of 75Ω ±5%. For receivers, video inputs should be AC-coupled and have a 75-Ω ±5% input impedance. The three signals must be coincident with respect to each other within ±5 ns. Sync information may be present on just the green channel, all three channels, as a separate composite sync signal, or as separate horizontal and vertical sync signals. A gamma of 1/0.45 is used.
73
As shown in Figure 5.5, the nominal active video amplitude is 700 mV, and has no blanking pedestal. A ±300 ±6 mV tri-level composite sync signal may be present on just the green channel (consumer market), or all three channels (pro-video market). DC offsets up to ±1V may be present.
Analog R´G´B´ Generation Assuming 10-bit DACs with an output range of 0–1.305V (to match the video DACs used by the NTSC/PAL encoder in Chapter 9), the 10-bit YCbCr to R´G´B´ equations are: R´ = 0.625(Y709 – 64) + 0.963(Cr – 512) G´ = 0.625(Y709 – 64) – 0.287(Cr – 512) – 0.114(Cb – 512) B´ = 0.625(Y709 – 64) + 1.136(Cb – 512)
R´G´B´ has a nominal 10-bit range of 0–548 to match the active video levels used by the NTSC/PAL encoder in Chapter 9. Note that negative values of R´G´B´ should be supported at this point. The R´G´B´ data is clamped by a blanking signal that has a raised cosine distribution to slow the slew rate of the start and end of the video signal. For 1080i and 720p systems, blank rise and fall times are 54 ±20 ns. For 1080p systems, blank rise and fall times are 27 ±10 ns. Composite sync information may be added to the R´G´B´ data after the blank processing has been performed. Values of 16 (sync low), 488 (high sync), or 252 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution to slow the slew rate of the sync signal. For 1080i systems, sync rise and fall times are 54 ±20 ns, and the horizontal sync low and high widths at the 50%-points are 593 ±40 ns. For
74
Chapter 5: Analog Video Interfaces
WHITE LEVEL (800)
1.020 V
100 IRE
0.622 V
SYNC LEVEL (488)
43 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
GREEN, BLUE, OR RED CHANNEL, SYNC PRESENT
WHITE LEVEL (800)
1.020 V
100 IRE
0.321 V
BLACK / BLANK LEVEL (252)
GREEN, BLUE, OR RED CHANNEL, NO SYNC PRESENT
Figure 5.5. HDTV Analog RGB Levels. 0 IRE blanking level.
SDTV YPbPr Interface
720p systems, sync rise and fall times are 54 ±20 ns, and the horizontal sync low and high widths at the 50%-points are 539 ±40 ns. For 1080p systems, sync rise and fall times are 27 ±10 ns, and the horizontal sync low and high widths at the 50%-points are 296 ±20 ns. At this point, we have digital R´G´B´ with sync and blanking information, as shown in Figure 5.5 and Table 5.3. The numbers in parentheses in Figure 5.5 indicate the data value for a 10-bit DAC with a full-scale output value of 1.305V. The digital R´G´B´ data drive three 10-bit DACs to generate the analog R´G´B´ video signals. Video Level
0 IRE Blanking Pedestal
white
800
sync - high
488
black
252
blank
252
sync - low
16
Table 5.3. HDTV 10-Bit R´G´B´ Values.
Analog R´G´B´ Digitization Assuming 10-bit ADCs with an input range of 0–1.305V (to match the video ADCs used by the NTSC/PAL decoder in Chapter 9), the 10bit R´G´B´ to YCbCr equations are: Y709 = 0.341(R´ – 252) + 1.143(G´ – 252) + 0.115(B´ – 252) + 64 Cb = –0.188(R´ – 252) – 0.629(G´ – 252) + 0.817(B´ – 252) + 512 Cr = 0.817(R´ – 252) – 0.743(G´ – 252) – 0.074(B´ – 252) + 512
75
R´G´B´ has a nominal 10-bit range of 252– 800 to match the active video levels used by the NTSC/PAL decoder in Chapter 9. Table 5.3 and Figure 5.5 illustrate the 10-bit R´G´B´ values for the white, black, blank, and (optional) sync levels.
Copy Protection Currently, there is no approved copy protection technology for this high-definition analog interface. For this reason, some standards and DRM implementations disable these video outputs or only allow a “constrained image” to be output. A “constrained image” has an effective maximum resolution of 960 × 540p, although the total number of video samples and the video timing remain unchanged (for example, 1280 × 720p or 1920 × 1080i).
SDTV YPbPr Interface Some SDTV consumer video equipment supports an analog YPbPr video interface. NTSC and PAL VBI (vertical blanking interval) data, discussed in Chapter 8, may be present on 480i or 576i Y video signals. Three separate RCA phono connectors (consumer market) or BNC connectors (pro-video market) are used. The horizontal and vertical video timing are dependent on the video standard, as discussed in Chapter 4. For sources, the video signal at the connector should have a source impedance of 75Ω ±5%. For receivers, video inputs should be AC-coupled and have a 75-Ω ±5% input impedance. The three signals must be coincident with respect to each other within ±5 ns.
76
Chapter 5: Analog Video Interfaces
WHITE LEVEL (800)
1.020 V
100 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
Y CHANNEL, SYNC PRESENT
PEAK LEVEL (786)
1.003 V
50 IRE
0.653 V
BLACK / BLANK LEVEL (512)
50 IRE
0.303 V
PEAK LEVEL (238)
PB OR PR CHANNEL, NO SYNC PRESENT
Figure 5.6. EIA-770.2 SDTV Analog YPbPr Levels. Sync on Y.
SDTV YPbPr Interface
WHITE LEVEL (800)
1.020 V
100 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
Y CHANNEL, SYNC PRESENT
PEAK LEVEL (786)
1.003 V
50 IRE
0.653 V
BLACK / BLANK LEVEL (512)
50 IRE
43 IRE
0.352 V
SYNC LEVEL (276)
0.303 V
PEAK LEVEL (238)
PB OR PR CHANNEL, SYNC PRESENT
Figure 5.7. SDTV Analog YPbPr Levels. Sync on YPbPr.
77
White
Yellow
Cyan
Green
Magenta
Red
Blue
Black
Chapter 5: Analog Video Interfaces
IRE
100
88.6
70.1
58.7
41.3
29.9
11.4
0
mV
700
620
491
411
289
209
80
0
IRE
0
–50
16.8
–33.1
33.1
–16.8
50
0
mV
0
–350
118
–232
232
–118
350
0
IRE
0
8.1
–50
–41.8
41.8
50
–8.1
0
mV
0
57
–350
–293
293
350
–57
0
Y
64 to 940
940
840
678
578
426
326
164
64
Cb
64 to 960
512
64
663
215
809
361
960
512
Cr
64 to 960
512
585
64
137
887
960
439
512
Y
Pb
Pr
Pb
Pr
66.5
52.6
44
31
Black
Magenta
75
Blue
Green
IRE
Red
Cyan
Y
Yellow
Table 5.4. EIA-770.2 SDTV YPbPr and YCbCr 100% Color Bars. YPbPr values relative to the blanking level.
White
78
22.4
8.6
0
mV
525
465
368
308
217
157
60
0
IRE
0
–37.5
12.6
–24.9
24.9
–12.6
37.5
0
mV
0
–262
88
–174
174
–88
262
0
IRE
0
6.1
–37.5
–31.4
31.4
37.5
–6.1
0
mV
0
43
–262
–220
220
262
–43
0
Y
64 to 940
721
646
525
450
335
260
139
64
Cb
64 to 960
512
176
625
289
735
399
848
512
Cr
64 to 960
512
567
176
231
793
848
457
512
Table 5.5. EIA-770.2 SDTV YPbPr and YCbCr 75% Color Bars. YPbPr values relative to the blanking level.
SDTV YPbPr Interface
For consumer products, composite sync is present on only the Y channel. For pro-video applications, composite sync is present on all three channels. A gamma of 1/0.45 is specified. As shown in Figures 5.6 and 5.7, the Y signal consists of 700 mV of active video (with no blanking pedestal). Pb and Pr have a peak-topeak amplitude of 700 mV. A 300 ±6 mV composite sync signal is present on just the Y channel (consumer market), or all three channels (pro-video market). DC offsets up to ±1V may be present. The 100% and 75% YPbPr color bar values are shown in Tables 5.4 and 5.5.
Analog YPbPr Generation Assuming 10-bit DACs with an output range of 0–1.305V (to match the video DACs used by the NTSC/PAL encoder in Chapter 9), the 10-bit YCbCr to YPbPr equations are: Y = ((800 – 252)/(940 – 64))(Y601 – 64) Pb = ((800 – 252)/(960 – 64))(Cb – 512) Pr = ((800 – 252)/(960 – 64))(Cr – 512)
Y has a nominal 10-bit range of 0–548 to match the active video levels used by the NTSC/PAL encoder in Chapter 9. Pb and Pr have a nominal 10-bit range of 0 to ±274. Note that negative values of Y should be supported at this point. The YPbPr data is clamped by a blanking signal that has a raised cosine distribution to slow the slew rate of the start and end of the video signal. For 480i and 576i systems, blank rise and fall times are 140 ±20 ns. For 480p and 576p systems, blank rise and fall times are 70 ±10 ns.
79
Composite sync information is added to the Y data after the blank processing has been performed. Values of 16 (sync present) or 252 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution (between 16 and 252) to slow the slew rate of the sync signal. Composite sync information may also be added to the PbPr data after the blank processing has been performed. Values of 276 (sync present) or 512 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution (between 276 and 512) to slow the slew rate of the sync signal. For 480i and 576i systems, sync rise and fall times are 140 ±20 ns, and horizontal sync width at the 50%-point is 4.7 ±0.1 µs. For 480p and 576p systems, sync rise and fall times are 70 ±10 ns, and horizontal sync width at the 50%point is 2.33 ±0.05 µs. At this point, we have digital YPbPr with sync and blanking information, as shown in Figures 5.6 and 5.7 and Table 5.6. The numbers in parentheses in Figures 5.6 and 5.7 indicate the data value for a 10-bit DAC with a fullscale output value of 1.305V. The digital YPbPr data drive three 10-bit DACs to generate the analog YPbPr video signals. Video Level
Y
PbPr
white
800
512
black
252
512
blank
252
512
sync
16
276
Table 5.6. SDTV 10-Bit YPbPr Values.
80
Chapter 5: Analog Video Interfaces
Analog YPbPr Digitization Assuming 10-bit ADCs with an input range of 0–1.305V (to match the video ADCs used by the NTSC/PAL decoder in Chapter 9), the 10bit YPbPr to YCbCr equations are: Y601 = 1.5985(Y – 252) + 64 Cb = 1.635(Pb – 512) + 512 Cr = 1.635(Pr – 512) + 512
Y has a nominal 10-bit range of 252–800 to match the active video levels used by the NTSC/PAL decoder in Chapter 9. Table 5.6 and Figures 5.6 and 5.7 illustrate the 10-bit YPbPr values for the white, black, blank, and (optional) sync levels.
VBI Data for 480p Systems CGMS EIA/CEA-805, IEC 61880–2 and EIA-J CPR–1204–1 define the format of CGMS (Copy Generation Management System) data on line
70 ±10 IRE
41 for 480p systems. The waveform is illustrated in Figure 5.8. A sample clock rate of 27 MHz (59.94 Hz frame rate) or 27.027 MHz (60 Hz frame rate) is used. Each data bit is 26 clock cycles, or 963 ±30 ns, wide with a maximum rise and fall time of 50 ns. A logical “1” has an amplitude of 70 ±10 IRE; a logical “0” has an amplitude of 0 ±5 IRE. The 2-bit start symbol begins 156 clock cycles, or about 5.778 µs, after 0H. It consists of a “1” followed by a “0.” The 6-bit header follows the start symbol, and defines the nature of the payload data as shown in Table 5.7. The End of Message immediately follows the last packet of any data service that uses more than one packet. It has an associated payload consisting of all zeros. ECCI is a data service that may use more than one packet, thus requiring the use of the End of Message. The 14-bit payload for CGMS data is shown in Table 5.8. The 14-bit payload for ECCI data is currently “reserved”, consisting of all ones.
START SYMBOL
HEADER (H0 - H5)
DATA (D0 - D13)
2 BITS
6 BITS
14 BITS
BLANK LEVEL
SYNC LEVEL
5.778 µS
Figure 5.8. EIA-805, IEC 61880–2 and EIA-J CPR–1204–1 480p Line 41 CGMS Timing.
SDTV YPbPr Interface
H0
H1
Aspect Ratio
Picture Display Format
0
0
4:3
normal
0
1
16:9
normal
1
0
4:3
letter box
1
1
reserved
reserved
H2
H3
H4
H5
0
0
0
0
CGMS
0
0
0
1
Extended Copy Control Information (ECCI)
0
0
1
0
:
Service Name
reserved
1
1
1
0
1
1
1
1
End of Message (default if no copyright information)
Table 5.7. EIA-805, IEC 61880–2 and EIA-J CPR–1204–1 Header Format.
81
82
Chapter 5: Analog Video Interfaces
D0
D1
CGMS
CGMS
D2
D3
D4
APS APS ASB
D5
D6
D7
0
0
0
D8
D9
D10 D11 D12 D13
CRC = x6 + x + 1
D0–D1: CGMS Definition 00 copying permitted without restriction 01 no more copies (one copy has already been made) 10 one copy allowed 11 no copying permitted D2–D3: Analog Protection Services (valid only if both D0 and D1 are “1”) 00 no Macrovision pseudo-sync pulse 01 Macrovision pseudo-sync pulse on; color striping off 10 Macrovision pseudo-sync pulse on; 2-line color striping on 11 Macrovision pseudo-sync pulse on; 4-line color striping on D4: Analog Source Bit 0 non-analog pre-recorded package medium 1 analog pre-recorded package medium
Table 5.8. EIA-805, IEC 61880–2 and EIA-J CPR–1204–1 CGMS Payload Format.
SDTV YPbPr Interface
VBI Data for 576p Systems CGMS IEC 62375 defines the format of CGMS (Copy Generation Management System) and widescreen signalling (WSS) data on line 43 for 576p systems. The waveform is illustrated in Figure 5.9. This standard allows a WSSenhanced 16:9 TV to display programs in their correct aspect ratio. Data Timing CGMS and WSS data is normally on line 43, as shown in Figure 5.9. However, due to video editing, the data may reside on any line between 43–47. The clock frequency is 10 MHz (±1 kHz). The signal waveform should be a sine-squared pulse, with a half-amplitude duration of 100 ±10 ns. The signal amplitude is 500 mV ±5%.
500 MV ±5%
The NRZ data bits are processed by a biphase code modulator, such that one data period equals 6 elements at 10 MHz. Data Content The WSS consists of a run-in code, a start code, and 14 bits of data, as shown in Table 5.9. Run-In The run-in consists of 29 elements of a specific sequence at 10 MHz, shown in Table 5.9. Start Code The start code consists of 24 elements of a specific sequence at 10 MHz, shown in Table 5.9.
RUN IN
START CODE
DATA (B0 - B13)
29 10 MHZ ELEMENTS
24 10 MHZ ELEMENTS
84 10 MHZ ELEMENTS
BLANK LEVEL
43 IRE
SYNC LEVEL
5.5 ± 0.125 µS
83
13.7 µS
FIgure 5.9. IEC 62375 576p Line 43 CGMS Timing.
90–110 NS RISE / FALL TIMES (2T BAR SHAPING)
84
Chapter 5: Analog Video Interfaces
Group 1 Data The group 1 data consists of 4 data bits that specify the aspect ratio. Each data bit generates 6 elements at 10 MHz. b0 is the LSB. Table 5.9 lists the data bit assignments and usage. The number of active lines listed in Table 5.10 are for the exact aspect ratio (a = 1.33, 1.56, or 1.78). The aspect ratio label indicates a range of possible aspect ratios (a) and number of active lines: 4:3 14:9 16:9 >16:9
a ≤ 1.46 1.46 < a ≤ 1.66 1.66 < a ≤ 1.90 a > 1.90
527–576 463–526 405–462 < 405
To allow automatic selection of the display mode, a 16:9 receiver should support the following minimum requirements:
Group 3 Data The group 3 data consists of three data bits that specify subtitles. Each data bit generates six elements at 10 MHz. Data bit b8 is the LSB. b10, b9: open subtitles 00 no 01 inside active picture 10 outside active picture 11 reserved
Group 4 Data The group 4 data consists of three data bits that specify surround sound and copy protection. Each data bit generates six elements at 10 MHz. Data bit b11 is the LSB. b11: surround sound 0 no 1 yes
Case 1: The 4:3 aspect ratio picture should be centered on the display, with black bars on the left and right sides.
b12: copyright 0 no copyright asserted or unknown 1 copyright asserted
Case 2: The 14:9 aspect ratio picture should be centered on the display, with black bars on the left and right sides. Alternately, the picture may be displayed using the full display width by using a small (typically 8%) horizontal geometrical error.
b13: copy protection 0 copying not restricted 1 copying restricted
Case 3: The 16:9 aspect ratio picture should be displayed using the full width of the display. Case 4: The >16:9 aspect ratio picture should be displayed as in Case 3 or use the full height of the display by zooming in.
SDTV YPbPr Interface
run-in
29 elements at 10 MHz
1 1111 0001 1100 0111 0001 1100 0111 (1F1C 71C7H)
start code
24 elements at 10 MHz
0001 1110 0011 1100 0001 1111 (1E 3C1FH)
group 1 (aspect ratio)
group 2 (enhanced services)
group 3 (subtitles)
group 4 (reserved)
24 elements at 10 MHz “0” = 000 111 “1” = 111 000 24 elements at 10 MHz “0” = 000 111 “1” = 111 000 18 elements at 10 MHz “0” = 000 111 “1” = 111 000 18 elements at 10 MHz “0” = 000 111 “1” = 111 000
b0, b1, b2, b3
b4, b5, b6, b7 (b4, b5, b6 and b7 = “0” since reserved)
b8, b9, b10 (b8 = “0” since reserved)
b11, b12, b13
Table 5.9. IEC 62375 576p Line 43 WSS Information.
b3, b2, b1, b0
Aspect Ratio Label
Format
Position On 4:3 Display
Active Lines
Minimum Requirements
1000
4:3
full format
–
576
case 1
0001
14:9
letterbox
center
504
case 2
0010
14:9
letterbox
top
504
case 2
1011
16:9
letterbox
center
430
case 3
0100
16:9
letterbox
top
430
case 3
1101
> 16:9
letterbox
center
–
case 4
1110
14:9
full format
center
576
–
0111
16:9
full format (anamorphic)
–
576
–
Table 5.10. IEC 62375 Group 1 (Aspect Ratio) Data Bit Assignments and Usage.
85
86
Chapter 5: Analog Video Interfaces
HDTV YPbPr Interface Some HDTV consumer video equipment supports an analog YPbPr video interface. Three separate RCA phono connectors (consumer market) or BNC connectors (pro-video market) are used. The horizontal and vertical video timing are dependent on the video standard, as discussed in Chapter 4. For sources, the video signal at the connector should have a source impedance of 75Ω ±5%. For receivers, video inputs should be AC-coupled and have a 75-Ω ±5% input impedance. The three signals must be coincident with respect to each other within ±5 ns. For consumer products, composite sync is present on only the Y channel. For pro-video applications, composite sync is present on all three channels. A gamma of 1/0.45 is specified. As shown in Figures 5.10 and 5.11, the Y signal consists of 700 mV of active video (with no blanking pedestal). Pb and Pr have a peakto-peak amplitude of 700 mV. A ±300 ±6 mV composite sync signal is present on just the Y channel (consumer market), or all three channels (pro-video market). DC offsets up to ±1V may be present. The 100% and 75% YPbPr color bar values are shown in Tables 5.11 and 5.12.
Analog YPbPr Generation Assuming 10-bit DACs with an output range of 0–1.305V (to match the video DACs used by the NTSC/PAL encoder in Chapter 9), the 10-bit YCbCr to YPbPr equations are: Y = ((800 – 252)/(940 – 64))(Y709 – 64) Pb = ((800 – 252)/(960 – 64))(Cb – 512) Pr = ((800 – 252)/(960 – 64))(Cr – 512)
Y has a nominal 10-bit range of 0–548 to match the active video levels used by the NTSC/PAL encoder in Chapter 9. Pb and Pr have a nominal 10-bit range of 0 to ±274. Note that negative values of Y should be supported at this point. The YPbPr data is clamped by a blanking signal that has a raised cosine distribution to slow the slew rate of the start and end of the video signal. For 1080i and 720p systems, blank rise and fall times are 54 ±20 ns. For 1080p systems, blank rise and fall times are 27 ±10 ns. Composite sync information is added to the Y data after the blank processing has been performed. Values of 16 (sync low), 488 (high sync), or 252 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution to slow the slew rate of the sync signal. Composite sync information may be added to the PbPr data after the blank processing has been performed. Values of 276 (sync low), 748 (high sync), or 512 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution to slow the slew rate of the sync signal. For 1080i systems, sync rise and fall times are 54 ±20 ns, and the horizontal sync low and high widths at the 50%-points are 593 ±40 ns. For 720p systems, sync rise and fall times are 54 ±20 ns, and the horizontal sync low and high widths at the 50%-points are 539 ±40 ns. For 1080p systems, sync rise and fall times are 27 ±10 ns, and the horizontal sync low and high widths at the 50%-points are 296 ±20 ns. At this point, we have digital YPbPr with sync and blanking information, as shown in Figures 5.10 and 5.11 and Table 5.13. The numbers in parentheses in Figures 5.10 and 5.11 indicate the data value for a 10-bit DAC with a full-scale output value of 1.305V. The digital
HDTV YPbPr Interface
WHITE LEVEL (800)
1.020 V
100 IRE 0.622 V
SYNC LEVEL (488)
43 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
Y CHANNEL, SYNC PRESENT
PEAK LEVEL (786)
1.003 V
50 IRE
0.653 V
BLACK / BLANK LEVEL (512)
50 IRE
0.303 V
PEAK LEVEL (238)
PB OR PR CHANNEL, NO SYNC PRESENT
Figure 5.10. EIA-770.3 HDTV Analog YPbPr Levels. Sync on Y.
87
88
Chapter 5: Analog Video Interfaces
WHITE LEVEL (800)
1.020 V
100 IRE 0.622 V
SYNC LEVEL (488)
43 IRE
0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
Y CHANNEL, SYNC PRESENT
1.003 V
PEAK LEVEL (786)
0.954 V
SYNC LEVEL (748)
50 IRE 43 IRE
0.653 V
BLACK / BLANK LEVEL (512)
50 IRE
43 IRE
0.352 V
SYNC LEVEL (276)
0.303 V
PEAK LEVEL (238)
PB OR PR CHANNEL, SYNC PRESENT
Figure 5.11. SMPTE 274M and 296M HDTV Analog YPbPr Levels. Sync on YPbPr.
Yellow
Cyan
Green
Magenta
Red
Blue
Black
Y
White
HDTV YPbPr Interface
IRE
100
92.8
78.7
71.5
28.5
21.3
7.2
0
mV
700
650
551
501
200
149
50
0
IRE
0
–50
11.4
–38.5
38.5
–11.4
50
0
mV
0
–350
80
–270
270
–80
350
0
IRE
0
4.6
–50
–45.4
45.4
50
–4.6
0
mV
0
32
–350
–318
318
350
–32
0
Y
64 to 940
940
877
753
690
314
251
127
64
Cb
64 to 960
512
64
614
167
857
410
960
512
Cr
64 to 960
512
553
64
106
918
960
471
512
Pb
Pr
Pr
59
53.7
21.3
525
487
413
376
149
0
–37.5
8.6
–28.9
28.9
Black
69.6
Blue
Magenta
75
Red
Green
mV IRE
Cyan
Pb
IRE
Yellow
Y
White
Table 5.11. EIA-770.3 HDTV YPbPr and YCbCr 100% Color Bars. YPbPr values relative to the blanking level.
16
5.4
0
112
38
0
–8.6
37.5
0
mV
0
–263
60
–202
202
–60
263
0
IRE
0
3.5
–37.5
–34
34
37.5
–3.5
0
mV
0
24
–263
–238
238
263
–24
0
Y
64 to 940
721
674
581
534
251
204
111
64
Cb
64 to 960
512
176
589
253
771
435
848
512
Cr
64 to 960
512
543
176
207
817
848
481
512
Table 5.12. EIA-770.3 HDTV YPbPr and YCbCr 75% Color Bars. YPbPr values relative to the blanking level.
89
90
Chapter 5: Analog Video Interfaces
YPbPr data drive three 10-bit DACs to generate the analog YPbPr video signals. Video Level
Y
PbPr
white
800
512
sync - high
488
748
black
252
512
blank
252
512
sync - low
16
276
Table 5.13. HDTV 10-Bit YPbPr Values.
Analog YPbPr Digitization Assuming 10-bit ADCs with an input range of 0–1.305V (to match the video ADCs used by the NTSC/PAL decoder in Chapter 9), the 10bit YPbPr to YCbCr equations are: Y709 = 1.5985(Y – 252) + 64
VBI Data for 720p Systems CGMS EIA/CEA-805 and EIA-J CPR–1204–2 define the format of CGMS (Copy Generation Management System) data on line 24 for 720p systems. The waveform is illustrated in Figure 5.12. A sample clock rate of 74.176 MHz (59.94 Hz frame rate) or 74.25 MHz (60 Hz frame rate) is used. Each data bit is 58 clock cycles, or 782 ±30 ns, wide with a maximum rise and fall time of 50 ns. A logical “1” has an amplitude of 70 ±10 IRE; a logical “0” has an amplitude of 0 ±5 IRE. The 2-bit start symbol begins 232 clock cycles, or about 3.128 µs, after 0H. It consists of a “1” followed by a “0.” The 6-bit header and 14-bit CGMS payload data format is the same as for 480p systems discussed earlier in this chapter.
VBI Data for 1080i Systems
Cb = 1.635(Pb – 512) + 512 Cr = 1.635(Pr – 512) + 512
Y has a nominal 10-bit range of 252–800 to match the active video levels used by the NTSC/PAL decoder in Chapter 9. Table 5.13 and Figures 5.10 and 5.11 illustrate the 10-bit YPbPr values for the white, black, blank, and (optional) sync levels.
CGMS EIA/CEA-805 and EIA-J CPR–1204–2 define the format of CGMS (Copy Generation Management System) data on lines 19 and 582 for 1080i systems. The waveform is illustrated in Figure 5.13. A sample clock rate of 74.176 MHz (59.94 Hz field rate) or 74.25 MHz (60 Hz field rate) is used. Each data bit is 77 clock cycles, or 1038 ±30 ns, wide with a maximum rise and fall time of 50 ns. A logical “1” has an amplitude of 70 ±10 IRE; a logical “0” has an amplitude of 0 ±5 IRE.
D-Connector Interface
The 2-bit start symbol begins 308 clock cycles, or about 4.152 µs, after 0H. It consists of a “1” followed by a “0.” The 6-bit header and 14-bit CGMS payload data format is the same as for 480p systems discussed earlier in this chapter.
Copy Protection Currently, there is no approved copy protection technology for this high-definition analog interface. For this reason, some standards and DRM implementations disable these video outputs or only allow a “constrained image” to be output. A “constrained image” has an effective maximum resolution of 960 × 540p, although the total number of video samples and the video timing remain unchanged (for example, 1280 × 720p or 1920 × 1080i).
70 ±10 IRE
91
D-Connector Interface A 14-pin female D-Connector (EIA-J CP– 4120 standard, EIA-J RC–5237 connector) is optionally used on some high-end consumer equipment in Japan, Hong Kong and Singapore. It is used to transfer EIA 770.2 or EIA 770.3 interlaced or progressive analog YPbPr video. There are five flavors of the D-Connector, referred to as D1, D2, D3, D4 and D5, each used to indicate supported video formats, as indicated in Table 5.14. Figure 5.14 illustrates the connector and Table 5.15 lists the pin names. Three Line signals (Line 1, Line 2 and Line 3) indicate the resolution and refresh rate of the YPbPr source video, as indicated in Table 5.16.
START SYMBOL
HEADER (H0 - H5)
DATA (D0 - D13)
2 BITS
6 BITS
14 BITS
BLANK LEVEL
SYNC LEVEL
3.128 µS
Figure 5.12. EIA-805 and EIA-J CPR–1204–2 720p Line 24 CGMS Timing.
92
Chapter 5: Analog Video Interfaces
Other Pro-Video Analog Interfaces
References
Tables 5.17 and 5.18 list some other common component analog video formats. The horizontal and vertical timing is the same as for 525-line (M) NTSC and 625-line (B, D, G, H, I) PAL. The 100% and 75% color bar values are shown in Tables 5.19 through 5.22. The SMPTE, EBU N10, 625-line Betacam, and 625line MII values are the same as for SDTV YPbPr.
VGA Interface Table 5.23 and Figure 5.15 illustrate the 15pin VGA connector used by computer equipment, and some consumer equipment, to transfer analog RGB signals. The analog RGB signals do not contain sync information and have no blanking pedestal, as shown in Figure 5.4.
70 ±10 IRE
1. EIA–770.1, Analog 525-Line Component Video Interface—Three Channels, November 2001. 2. EIA–770.2, Standard Definition TV Analog Component Video Interface, November 2001. 3. EIA–770.3, High Definition TV Analog Component Video Interface, November 2001. 4. EIA/CEA–805, Data Services on the Component Video Interfaces, October 2000. 5. EIA-J CPR–1204–1, Transfer Method of Video ID Information using Vertical Blanking Interval (525P System), March 1998. 6. EIA-J CPR–1204–2, Transfer Method of Video ID Information using Vertical Blanking Interval (720P, 1125I System), January 2000. 7. EIA-J CP–4120, Interface Between Digital Tuner and Television Receiver Using DConnector, January 2000.
START SYMBOL
HEADER (H0 - H5)
DATA (D0 - D13)
2 BITS
6 BITS
14 BITS
BLANK LEVEL
SYNC LEVEL
4.152 µS
Figure 5.13. EIA-805 and EIA-J CPR–1204–2 1080i Lines 19 and 582 CGMS Timing.
D1 D2 D3 D4 D5
× ×
×
×
×
× ×
× ×
× ×
1080p
1080i
720p
480p
480i
References
×
× ×
×
Table 5.14. D-Connector Supported Video Formats.
Figure 5.14. D-Connector.
Pin 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Function Y ground - Y Pb ground - Pb Pr ground - Pr reserved 1 line 1 line 2 reserved 2 line 3 ground - detect plugged reserved 3 detect plugged
Signal Level
Impedance
0.700V + sync
75 ohms
±0.350V ±0.350V
0V, 2.2V, or 5V1 0V, 2.2V, or 5V1
75 ohms 75 ohms 10K ±3Κ ohm 10K ±3Κ ohm
0V, 2.2V, or 5V1
10K ±3Κ ohm
0V = plugged in2
> 100K ohm
Notes: 1. 2.2V has range of 2.2V ±0.8V. 5V has a range of 5V ±1.5V. 2. Inside equipment, pin 12 is connected to ground and pin 14 is pulled to 5V through a resistor. Inside each D-Connector plug, pins 12 and 14 are shorted together.
Table 5.15. D-Connector Pin Descriptions.
93
94
Chapter 5: Analog Video Interfaces
Resolution
Refresh Rate
Line 1 Scan Lines
Line 2 Refresh Rate
Line 3 Aspect Ratio
60i
5V
0V
5V
2
5V
2.2V
5V
30p
5V
2.2V
5V
25p2
5V
2.2V
5V
24p2
5V
2.2V
5V
24sF2
5V
2.2V
5V
50i 1920x1080
60p
2.2V
5V
5V
50p2
2.2V
2.2V
5V
30p
2.2V
2.2V
5V
25p2
2.2V
2.2V
5V
24p2
2.2V
2.2V
5V
60p2
0V
5V
0V
16:9 Squeeze
60p
0V
5V
5V
16:9 Squeeze
60i
0V
0V
5V
16:9 Letterbox
60i
0V
0V
2.2V
4:3
60i
0V
0V
0V
1280x720
640x480
720x480
Chromaticity and Reference White
Color Sync Gamma Space Amplitude Correction Equations on Y
EIA-770.3
EIA-770.3
EIA-770.3
±0.300V3
EIA-770.2
EIA-770.2
EIA-770.2
–0.300V3
Notes: 1. 60p, 60i, 30p, and 24p refresh rates also include the 59.94p, 59.94i, 29.97p and 23.976p refresh rates. 2. Not part of EIAJ CP-4120 specification, but commonly supported by equipment. 3. Relative to the blanking level.
Table 5.16. Voltage Levels of Line Signals for Various Video Formats for D-Connector.
References
Format
SMPTE, EBU N10
525-Line Betacam1
625-Line Betacam1
525-Line MII2
625-Line MII2
Output Signal
Signal Amplitudes (volts)
Y
+0.700
sync
–0.300
R´–Y, B´–Y
±0.350
Y
+0.714
sync
–0.286
R´–Y, B´–Y
±0.467
Y
+0.700
sync
–0.300
R´–Y, B´–Y
±0.350
Y
+0.700
sync
–0.300
R´–Y, B´–Y
±0.324
Y
+0.700
sync
–0.300
R´–Y, B´–Y
±0.350
Notes
0% setup on Y 100% saturation three wire = (Y + sync), (R´–Y), (B´–Y) 7.5% setup on Y only 100% saturation three wire = (Y + sync), (R´–Y), (B´–Y) 0% setup on Y 100% saturation three wire = (Y + sync), (R´–Y), (B´–Y) 7.5% setup on Y only 100% saturation three wire = (Y + sync), (R´–Y), (B´–Y) 0% setup on Y 100% saturation three wire = (Y + sync), (R´–Y), (B´–Y)
Notes: 1. Trademark of Sony Corporation. 2. Trademark of Matsushita Corporation.
Table 5.17. Common Pro-Video Component Analog Video Formats.
95
96
Chapter 5: Analog Video Interfaces
Output Signal
Signal Amplitudes (volts)
SMPTE, EBU N10
G´, B´, R´
+0.700
sync
–0.300
NTSC (setup)
G´, B´, R´
+0.714
sync
–0.286
NTSC (no setup)
G´, B´, R´
+0.714
sync
–0.286
G´, B´, R´
+0.700
sync
–0.300
Format
MII1
Notes 0% setup on G´, B´, and R´ 100% saturation three wire = (G´ + sync), B´, R´ 7.5% setup on G´, B´, and R´ 100% saturation three wire = (G´ + sync), B´, R´ 0% setup on G´, B´, and R´ 100% saturation three wire = (G´ + sync), B´, R´ 7.5% setup on G´, B´, and R´ 100% saturation three wire = (G´ + sync), B´, R´
Notes: 1. Trademark of Matsushita Corporation.
Table 5.18. Common Pro-Video RGB Analog Video Formats.
Green
Magenta
Red
Blue
Black
R´–Y
Cyan
B´–Y
Yellow
Y
White
References
IRE
100
89.5
72.3
61.8
45.7
35.2
18.0
7.5
mV
714
639
517
441
326
251
129
54
IRE
0
–65.3
22.0
–43.3
43.3
–22.0
65.3
0
mV
0
–466
157
–309
309
–157
466
0
IRE
0
10.6
–65.3
–54.7
54.7
65.3
–10.6
0
mV
0
76
–466
–391
391
466
–76
0
Green
Magenta
Red
Blue
Black
R´–Y
Cyan
B´–Y
Yellow
Y
White
Table 5.19. 525-Line Betacam 100% Color Bars. Values are relative to the blanking level.
IRE
76.9
69.0
56.1
48.2
36.2
28.2
15.4
7.5
mV
549
492
401
344
258
202
110
54
IRE
0
–49.0
16.5
–32.5
32.5
–16.5
49.0
0
mV
0
–350
118
–232
232
–118
350
0
IRE
0
8.0
–49.0
–41.0
41.0
49.0
–8.0
0
mV
0
57
–350
–293
293
350
–57
0
Table 5.20. 525-Line Betacam 75% Color Bars. Values are relative to the blanking level.
97
Green
Magenta
Red
Blue
Black
R´–Y
Cyan
B´–Y
Yellow
Y
White
Chapter 5: Analog Video Interfaces
IRE
100
89.5
72.3
61.8
45.7
35.2
18.0
7.5
mV
700
626
506
433
320
246
126
53
IRE
0
–46.3
15.6
–30.6
30.6
–15.6
46.3
0
mV
0
–324
109
–214
214
–109
324
0
IRE
0
7.5
–46.3
–38.7
38.7
46.3
–7.5
0
mV
0
53
–324
–271
271
324
–53
0
Magenta
Red
Blue
Black
R´–Y
Green
B´–Y
Cyan
Y
Yellow
Table 5.21. 525-Line MII 100% Color Bars. Values are relative to the blanking level.
White
98
IRE
76.9
69.0
56.1
48.2
36.2
28.2
15.4
7.5
mV
538
483
393
338
253
198
108
53
IRE
0
–34.7
11.7
–23.0
23.0
–11.7
34.7
0
mV
0
–243
82
–161
161
–82
243
0
IRE
0
5.6
–34.7
–29.0
29.0
34.7
–5.6
0
mV
0
39
–243
–203
203
243
–39
0
Table 5.22. 525-Line MII 75% Color Bars. Values are relative to the blanking level.
5 10 15
1 6 11
FIgure 5.15. VGA 15-Pin D-SUB Female Connector.
References
Pin 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Function red green blue ground ground ground - red ground - green ground - blue +5V DC ground - HSYNC ground - VSYNC DDC SDA (data) HSYNC (horizontal sync) VSYNC (vertical sync) DDC SCL (clock)
Signal Level
Impedance
0.7v 0.7v 0.7v
75 ohms 75 ohms 75 ohms
99
≥ 2.4v ≥ 2.4v ≥ 2.4v ≥ 2.4v
Notes: 1. DDC = Display Data Channel.
Table 5.23. VGA Connector Signals.
8. IEC 60993–1, Audio, Video and Audiovisual Systems—Interconnections and Matching Values—Part 1: 21-pin Connector for Video Systems, Application No. 1, April 1988. 9. IEC 61880–2, Video Systems (525/60)— Video and Accompanied Data Using the Vertical Blanking Interval—Part 2: 525 Progressive Scan System, September 2002. 10. IEC 62375, Video Systems (625/50 Progressive)—Video and Accompanied Data Using the Vertical Blanking Interval—Analog Interface, February 2004. 11. ITU-R BT.709–5, 2002, Parameter Values for the HDTV Standards for Production and International Programme Exchange.
12. SMPTE 253M–1998, Television—ThreeChannel RGB Analog Video Interface. 13. SMPTE 274M–2003, Television—1920 × 1080 Image Sample Structure, Digital Representation and Digital Timing Reference Sequences for Multiple Picture Rates. 14. SMPTE 293M–2003, Television—720 × 483 Active Line at 59.94-Hz Progressive Scan Production—Digital Representation. 15. SMPTE RP-160–1997, Three-Channel Parallel Analog Component High-Definition Video Interface.
100
Chapter 6: Digital Video Interfaces
Chapter 6: Digital Video Interfaces Chapter 6
Digital Video Interfaces Pro-Video Component Interfaces Pro-video equipment, such as that used within studios, has unique requirements and therefore its own set of digital video interconnect standards. Table 6.1 lists the various provideo parallel and serial digital interface standards.
Video Timing Rather than digitize and transmit the blanking intervals, special sequences are inserted into the digital video stream to indicate the start of active video (SAV) and end of active video (EAV). These EAV and SAV sequences indicate when horizontal and vertical blanking are present and which field is being transmitted. They also enable the transmission of ancillary data such as digital audio, teletext, captioning, etc. during the blanking intervals.
100
The EAV and SAV sequences must have priority over active video data or ancillary data to ensure that correct video timing is always maintained at the receiver. The receiver decodes the EAV and SAV sequences to recover the video timing. The video timing sequence of the encoder is controlled by three timing signals discussed in Chapter 4: H (horizontal blanking), V (vertical blanking), and F (Field 1 or Field 2). A zero-to-one transition of H triggers an EAV sequence while a one-to-zero transition triggers an SAV sequence. F and V are allowed to change only at EAV sequences. Usually, both 8-bit and 10-bit interfaces are supported, with the 10-bit interface used to transmit 2 bits of fractional video data to minimize cumulative processing errors and to support 10-bit ancillary data. YCbCr or R´G´B´ data may not use the 10bit values of 000H–003H and 3FCH–3FFH, or the 8-bit values of 00H and FFH, since they are used for timing information.
Pro-Video Component Interfaces
Active Resolution (H × V)
Total Resolution1 (H × V)
Display Aspect Ratio
Frame Rate (Hz)
1× Y Sample Rate (MHz)
SDTV or HDTV
Digital Parallel Standard
Digital Serial Standard
720 × 480i
858 × 525i
4:3
29.97
13.5
SDTV
BT.656 BT.799 SMPTE 125M
BT.656 BT.799
720 × 480p
858 × 525p
4:3
59.94
27
SDTV
–
BT.1362 SMPTE 294M
720 × 576i
864 × 625i
4:3
25
13.5
SDTV
BT.656 BT.799
BT.656 BT.799
720 × 576p
864 × 625p
4:3
50
27
SDTV
–
BT.1362
960 × 480i
1144 × 525i
16:9
29.97
18
SDTV
BT.1302 BT.1303 SMPTE 267M
BT.1302 BT.1303
960 × 576i
1152 × 625i
16:9
25
18
SDTV
BT.1302 BT.1303
BT.1302 BT.1303
1280 × 720p
1650 × 750p
16:9
59.94
74.176
HDTV
SMPTE 274M
–
16:9
60
74.25
HDTV
SMPTE 274M
–
1280 × 720p
1650 × 750p
1920 × 1080i
2200 × 1125i
16:9
29.97
74.176
HDTV
BT.1120 SMPTE 274M
BT.1120 SMPTE 292M
1920 × 1080i
2200 × 1125i
16:9
30
74.25
HDTV
BT.1120 SMPTE 274M
BT.1120 SMPTE 292M
1920 × 1080p 2200 × 1125p
16:9
59.94
148.35
HDTV
BT.1120 SMPTE 274M
–
1920 × 1080p 2200 × 1125p
16:9
60
148.5
HDTV
BT.1120 SMPTE 274M
–
1920 × 1080i
2376 × 1250i
16:9
25
74.25
HDTV
BT.1120
BT.1120
1920 × 1080p 2376 × 1250p
16:9
50
148.5
HDTV
BT.1120
–
Table 6.1. Pro-Video Parallel and Serial Digital Interface Standards for Various Component Video Formats. 1i = interlaced, p = progressive.
101
102
Chapter 6: Digital Video Interfaces
The EAV and SAV sequences are shown in Table 6.2. The status word is defined as: F = “0” for Field 1 F = “1” for Field 2 V = “1” during vertical blanking H = “0” at SAV H = “1” at EAV P3–P0 = protection bits
preamble. The status word (optionally error corrected at the receiver, see Table 6.3) is used to recover the H, V, and F timing signals.
Ancillary Data
P3 = V ⊕ H P2 = F ⊕ H P1 = F ⊕ V P0 = F ⊕ V ⊕ H
where ⊕ represents the exclusive-OR function. These protection bits enable one- and two-bit errors to be detected and one-bit errors to be corrected at the receiver. For most progressive video systems, F is usually a “0” since there is no field information. For 4:2:2 YCbCr data, after each SAV sequence, the stream of active data words always begins with a Cb sample, as shown in Figure 6.1. In the multiplexed sequence, the co-sited samples (those that correspond to the same point on the picture) are grouped as Cb, Y, Cr. During blanking intervals, unless ancillary data is present, 10-bit Y or R´G´B´ values should be set to 040H and 10-bit CbCr values should be set to 200H. The receiver detects the EAV and SAV sequences by looking for the 8-bit FFH 00H 00H
General Format Ancillary data packets are used to transmit information (such as digital audio, closed captioning, and teletext data) during the blanking intervals. ITU-R BT.1364 and SMPTE 291M describe the ancillary data formats. During horizontal blanking, ancillary data may be transmitted in the interval between the EAV and SAV sequences. During vertical blanking, ancillary data may be transmitted in the interval between the SAV and EAV sequences. Multiple ancillary packets may be present in a horizontal or vertical blanking interval, but they must be contiguous with each other. Ancillary data should not be present where indicated in Table 6.4 as these regions may be affected by video switching. There are two types of ancillary data formats. The older Type 1 format uses a single data ID word to indicate the type of ancillary data; the newer Type 2 format uses two words for the data ID. The general packet format is shown in Table 6.5.
8-bit Data
preamble
status word
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
F
V
H
P3
P2
P1
P0
0
0
Table 6.2. EAV and SAV Sequence.
103
Pro-Video Component Interfaces
Received D5–D2
Received F, V, H (Bits D8–D6) 000
001
010
011
100
0000
000
000
0001
000
*
0010
000
0011
*
0100
000
101
110
111
000
*
000
*
*
111
*
111
*
111
111
111
*
*
011
*
101
*
*
*
010
*
100
*
*
111
*
*
011
*
*
110
*
0101
*
001
*
*
100
*
*
111
0110
*
011
011
011
100
*
*
011
0111
100
*
*
011
100
100
100
*
1000
000
*
*
*
*
101
110
*
1001
*
001
010
*
*
*
*
111
1010
*
101
010
*
101
101
*
101
1011
010
*
010
010
*
101
010
*
1100
*
001
110
*
110
*
110
110
1101
001
001
*
001
*
001
110
*
1110
*
*
*
011
*
101
110
*
1111
*
001
010
*
100
*
*
*
Notes: * = uncorrectable error.
Table 6.3. SAV and EAV Error Correction at Decoder. BT.601 H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0 4
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
268 (280)
0 0 0
0 0 0 4
CO–SITED X Y Z
C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
C R 2
Y 3
C R 718
Y 719
3 F F
1440
1716 (1728)
Figure 6.1. BT.656 Parallel Interface Data For One Scan Line. 480i; 4:2:2 YCbCr; 720 active samples per line; 27 MHz clock; 10-bit system. The values for 576i systems are shown in parentheses.
BT.656 4:2:2 VIDEO
104
Chapter 6: Digital Video Interfaces
Video Standard
Line Numbers Affected
480i (525i)
10, 273
480p (525p)
10
576i (625i)
6, 319
576p (625ip
6
720p (750p)
7
1080i (1125i)
7, 569
1080p (1125p)
7
Table 6.4. Lines Affected by Switching.
Data ID (DID) DID indicates the type of data being sent. The assignment of most of the DID values is controlled by the ITU and SMPTE to ensure equipment compatibility. A few DID values are available that don’t require registration. Some DID values are listed in Table 6.6. Secondary ID (SDID, Type 2 Only) SDID is also part of the data ID for Type 2 ancillary formats. The assignment of most of the SDID values is also controlled by the ITU and SMPTE to ensure equipment compatibility. A few SDID values are available that don’t require registration. Some SDID values are listed in Table 6.6. Data Block Number (DBN, Type 1 Only) DBN is used to allow multiple ancillary packets (sharing the same DID) to be put back together at the receiver. This is the case when there are more than 255 user data words required to be transmitted, thus requiring
more than one ancillary packet to be used. The DBN value increments by one for each consecutive ancillary packet. Data Count (DC) DC specifies the number of user data words in the packet. In 8-bit applications, it specifies the six MSBs of an 8-bit value, so the number of user data words must be an integral number of four. User Data Words (UDW) Up to 255 user data words may be present in the packet. In 8-bit applications, the number of user data words must be an integral number of four. Padding words may be added to ensure an integral number of four user data words are present. User data may not use the 10-bit values of 000H–003H and 3FCH–3FFH, or the 8-bit values of 00H and FFH, since they are used for timing information.
Pro-Video Component Interfaces
8-bit Data
105
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
even parity
Value of 0000 0000 to 1111 1111
data block number or SDID
D8
even parity
Value of 0000 0000 to 1111 1111
data count (DC)
D8
even parity
Value of 0000 0000 to 1111 1111
ancillar y data flag (ADF)
Value of 00 0000 0100 to 11 1111 1011
user data word 0
: Value of 00 0000 0100 to 11 1111 1011
user data word N check sum
D8
Sum of D0–D8 of data ID through last user data word. Preset to all zeros; carry is ignored.
Table 6.5. Ancillary Data Packet General Format.
106
Chapter 6: Digital Video Interfaces
8-bit DID Type 2
Function
8-bit DID Type 1
00H
undefined
80H
01H–03H
reserved
81H–83H
04H, 08H, 0CH
8-bit applications
84H
Function marked for deletion reserved end marker
10H–3FH
reserved
85H–BFH
reserved
40H–5FH
user application
C0H–DFH
user application
60H
timecode
EOH–EBH
registered
61H
closed captioning
ECH
AES control packet, group 4
registered
EDH
AES control packet, group 3
EEH
AES control packet, group 2
EFH
AES control packet, group 1
F4H
error detection
F5H
longitudinal timecode
62H–7FH
8-bit SDID Type 2
Function
00H
undefined format
F8H
AES extended packet, group 4
x0H
8-bit applications
F9H
AES audio data, group 4
x4H
8-bit applications
FAH
AES extended packet, group 3
x8H
8-bit applications
FBH
AES audio data, group 3
xCH
8-bit applications
FCH
AES extended packet, group 2
unassigned
FDH
AES audio data, group 2
FEH
AES extended packet, group 1
FFH
AES audio data, group 1
all others
Table 6.6. DID and SDID Assignments.
Pro-Video Component Interfaces
Audio Sampling Rate (kHz)
Samples per Frame: 29.97 Hz Video
Samples per Frame: 25 Hz Video
Samples per Frame
Samples per Field 1
Samples per Field 2
Exceptions: Frame Number
48.0
8008 / 5
1602
1601
–
–
1920
44.1
147147 / 100
1472
1471
23 47 71
1471 1471 1471
1764
16016 / 15
1068
1067
4 8 12
1068 1068 1068
1280
32
Exceptions: Number of Samples
107
Table 6.7. Isochronous Audio Sample Rates.
Audio Format ITU-R BT.1305 and SMPTE 272M describe the transmission of digital audio as ancillary data. 2–16 channels of up to 24-bit digital audio are supported, with sample rates of 32–48 kHz. Table 6.7 lists the number of audio samples per video frame for various audio sample rates. Audio data of up to 20 bits per sample is transferred using the format in Table 6.8. “V” is the AES/EBU sample valid bit, “U” is the AES/EBU user bit, and ”C” is the AES/EBU audio channel status bit. “P” is an even parity bit for the 26 previous bits in the sample (excluding D9 in the first and second words of the audio sample). Audio is represented as two’s complement linear PCM data. To support 24-bit audio samples, extended data packets may be used to transfer the four auxiliary bits of the AES/EBU audio stream. Audio data is formatted as 1–4 groups, defined by [gr 1] and [gr 0], with each group having 1–4 channels of audio data, defined by [ch 1] and [ch 0].
Optional control packets may be used on lines 12 and 275 (480i systems) or lines 8 and 320 (576i systems) to specify the sample rate, delay relative to the video, etc. If present, it must be transmitted prior to any audio packets. If not transmitted, a default condition of 48 kHz isochronous audio is assumed. Timecode Format ITU-R BT.1366 and SMPTE RP-188 define the transmission of timecode using ancillary data for 480i, 576i, and 1080i systems. The ancillary packet format is shown in Table 6.9, and is used to convey longitudinal (LTC) or vertical interval timecode (VITC) information. For additional information on the timecode format, and the meaning of the flags in Table 6.9, see the timecode discussion in Chapter 8. VITC 1 and VITC 2 timecode packets for 480p systems use lines 27 and 28, respectively. LTC, VITC 1 and VITC 2 timecode packets for 1080i systems use lines 10, 9 and 571, respectively.
108
Chapter 6: Digital Video Interfaces
8-bit Data
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
even parity
1
1
1
1
1
gr 1
gr 0
1
data block number (DBN)
D8
even parity
Value of 0000 0000 to 1111 1111
data count (DC)
D8
even parity
Value of 0000 0000 to 1111 1111
D8
A5
A4
A3
A2
A1
A0
ch 1
ch 0
Z
D8
A14
A13
A12
A11
A10
A9
A8
A7
A6
D8
P
C
U
V
A19
A18
A17
A16
A15
ancillar y data flag (ADF)
audio sample 0
:
audio sample N
check sum
D8
A5
A4
A3
A2
A1
A0
ch 1
ch 0
Z
D8
A14
A13
A12
A11
A10
A9
A8
A7
A6
D8
P
C
U
V
A19
A18
A17
A16
A15
D8
Sum of D0–D8 of data ID through last audio sample word. Preset to all zeros; carry is ignored.
Table 6.8. Digital Audio Ancillary Data Packet Format.
Pro-Video Component Interfaces
8-bit Data
109
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
1
0
0
0
0
0
SDID
D8
EP
0
1
1
0
0
0
0
0
data count (DC)
D8
EP
0
0
0
1
0
0
0
0
D8
EP
units of frames
DBB10
0
0
0
D8
EP
user group 1
DBB11
0
0
0
D8
EP
DBB12
0
0
0
D8
EP
user group 2
DBB13
0
0
0
D8
EP
units of seconds
DBB14
0
0
0
D8
EP
user group 3
DBB15
0
0
0
D8
EP
DBB16
0
0
0
D8
EP
user group 4
DBB17
0
0
0
D8
EP
units of minutes
DBB20
0
0
0
D8
EP
user group 5
DBB21
0
0
0
D8
EP
DBB22
0
0
0
D8
EP
user group 6
DBB23
0
0
0
D8
EP
units of hours
DBB24
0
0
0
D8
EP
user group 7
DBB25
0
0
0
D8
EP
DBB26
0
0
0
D8
EP
DBB27
0
0
0
ancillar y data flag (ADF)
flag 2
flag 1
flag 3
tens of frames
tens of seconds
timecode data
check sum
flag 4
flag 6
tens of minutes
flag 5
tens of hours
user group 8
Sum of D0–D8 of data ID through last timecode data word. Preset to all zeros; carry is ignored.
D8
Notes: EP = even parity for D0–D7.
Table 6.9. Timecode Ancillary Data Packet Format.
110
Chapter 6: Digital Video Interfaces
DBB17
DBB16
DBB15
DBB14
DBB13
DBB12
DBB11
DBB10
0
0
0
0
0
0
0
0
LTC
0
0
0
0
0
0
0
1
VITC 1
0
0
0
0
0
0
1
0
VITC 2
0
0
0
0
0
0
1
1
:
Definition
user defined
0
0
0
0
0
1
0
1
0
0
0
0
0
1
1
0
film data block
0
0
0
0
0
1
1
1
production data block
0
0
0
0
1
0
0
0 locally generated time address and user data
: 0
1
1
1
1
1
0
0
0
1
1
1
1
1
0
1
video tape data block
0
1
1
1
1
1
1
0
film data block
0
1
1
1
1
1
1
1
production data block
1
0
0
0
0
0
0
0
: 1
1
1
1
reserved 1
1
1
1
Table 6.10. Binary Bit Group 1 Definitions.
Binary Bit Group 1 The eight bits that comprise binary bit group 1 (DBB10–DBB17) specify the type of timecode and user data, as shown in Table 6.10. Binary Bit Group 2 The eight bits that comprise binary bit group 2 (DBB20–DBB27) specify line numbering and status information.
DBB20–DBB24 specify the VITC line select as shown in Table 6.11. These convey the VITC line number location. If DBB25 is a “1,” when the timecode information is converted into an analog VITC signal on line N, it must also be repeated on line N + 2. If DBB26 is a “1,” a timecode error was received, and the transmitted timecode has been interpolated from a previous timecode.
Pro-Video Component Interfaces
480i Systems DBB24
DBB23
DBB22
DBB21
DBB20
0
0
1
1
0
0
1
0
1
0
576i Systems
VITC on Line N
VITC on Line N + 2
VITC on Line N
VITC on Line N + 2
0
–
–
6, 319
8, 321
1
1
–
–
7, 320
9, 322
0
0
0
–
–
8, 321
10, 323
1
0
0
1
–
–
9, 322
11, 324
0
1
0
1
0
10, 273
12, 275
10, 323
12, 325
0
1
0
1
1
11, 274
13, 276
11, 324
13, 326
0
1
1
0
0
12. 275
14, 277
12, 325
14, 327
0
1
1
0
1
13, 276
15, 278
13, 326
15, 328
0
1
1
1
0
14, 277
16, 279
14, 327
16, 329
0
1
1
1
1
15, 278
17, 280
15, 328
17, 330
1
0
0
0
0
16, 279
18, 281
16, 329
18, 331
1
0
0
0
1
17, 280
19, 282
17, 330
19, 332
1
0
0
1
0
18, 281
20, 283
18, 331
20, 333
1
0
0
1
1
19, 282
–
19, 332
21, 334
1
0
1
0
0
20, 283
–
20, 333
22, 335
1
0
1
0
1
–
–
21, 334
–
1
0
1
1
0
–
–
22, 335
–
Table 6.11. VITC Line Select Definitions for 480i and 576i Systems.
111
112
Chapter 6: Digital Video Interfaces
If DBB27 is a “0,” the user group bits are processed to compensate for any latency. If a “1,” the user bits are retransmitted with no delay compensation. User Group Bits 32 bits of user data may be transferred with each timecode packet. User data is organized as eight groups of four bits each, with the D7 bit being the MSB. For additional information on user bits, see the timecode discussion in Chapter 8. SMPTE 266M SMPTE 266M also defines a digital vertical interval timecode (DVITC) for 480i and 576i systems. It is an 8-bit digital representation of the analog VITC signal, transferred using the 8 MSBs. If the VITC is present, it is carried on the Y data channel in the active portion of lines 14 and 277 (and optionally lines 16 and 279) for 480i systems; lines 19 and 332 (and optionally lines 21 and 334) are used for 576i systems. The 90 bits of VITC information are carried by 675 consecutive Y samples. A 10-bit value of 040H represents a “0;” a 10-bit value of 300H represents a “1.” Unused Y samples have a value of 040H. Closed Captioning Format ITU-R BT.1619 and SMPTE 334M define the ancillary packet format for EIA-608 closed captioning, as shown in Table 6.12. The field bit is a “0” for Field 2 and a “1” for Field 1. The offset value is a 5-bit unsigned integer which represents the offset (in lines) of the data insertion line, relative to line 9 or 272 for 480i systems and line 5 or 318 for 576i systems.
ITU-R BT.1619 and SMPTE 334M also define the ancillary packet format for EIA-708 digital closed captioning, as shown in Table 6.13. The payload is the EIA-708 caption distribution packet (CDP), which has a variable length. Error Detection Checksum Format ITU-R BT.1304 and SMPTE RP-165 define a checksum for error detection. The ancillary packet format is shown in Table 6.14. For 13.5 MHz 480i systems, the ancillary packet occupies sample words 1689–1711 on lines 9 and 272 (sample words 2261–2283 for 18 MHz 480i systems). For 13.5 MHz 576i systems, the ancillary packet occupies sample words 1701–1723 on lines 5 and 318 (sample words 2277–2299 for 18 MHz 576i systems). Note that these locations are immediately prior to the SAV code words. Checksums Two checksums are provided: one for a field of active video data and one for a full field of data. Each checksum is a 16-bit value calculated as follows: CRC = x16 + x12 + x5 + x1
For the active CRC, the starting and ending samples for 13.5 MHz 480i systems are sample word 0 on lines 21 and 284 (start) and sample word 1439 on lines 262 and 525 (end). The starting and ending samples for 13.5 MHz 576i systems are sample word 0 on lines 24 and 336 (start) and sample word 1439 on lines 310 and 622 (end). For the field CRC, the starting and ending samples for 13.5 MHz 480i systems are sample word 1444 on lines 12 and 275 (start) and sam-
Pro-Video Component Interfaces
8-bit Data
113
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
1
0
0
0
0
1
SDID
D8
EP
0
0
0
0
0
0
1
0
data count (DC)
D8
EP
0
0
0
0
0
0
1
1
line
D8
EP
field
0
0
caption word 0
D8
EP
D07
D06
D05
D04
D03
D02
D01
D00
caption word 1
D8
EP
D17
D16
D15
D14
D13
D12
D11
D10
check sum
D8
ancillar y data flag (ADF)
offset
Sum of D0–D8 of data ID through last caption word. Preset to all zeros; carry is ignored.
Notes: EP = even parity for D0–D7.
Table 6.12. EIA-608 Closed Captioning Ancillary Data Packet Format.
ple word 1439 on lines 8 and 271 (end). The starting and ending samples for 13.5 MHz 576i systems are sample word 1444 on lines 8 and 321 (start) and sample word 1439 on lines 4 and 317 (end). Error Flags Error flags indicate the status of the previous field. edh (error detected here): A “1” indicates that a transmission error was detected since one or more ancillary packets did not match its checksum.
eda (error detected already): A “1” indicates a transmission error was detected at a prior point in the data path. A device that receives data with this flag set should forward the data with the flag set and the edh flag reset to “0” if no further errors are detected. idh (internal error detected here): A “1” indicates that an error unrelated to the transmission has been detected. ida (internal error status): A “1” indicates data was received from a device that does not support this error detection method.
114
Chapter 6: Digital Video Interfaces
8-bit Data
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
1
0
0
0
0
1
SDID
D8
EP
0
0
0
0
0
0
0
1
data count (DC)
D8
EP
Value of 0000 0000 to 1111 1111
data word 0
D8
EP
Value of 0000 0000 to 1111 1111
ancillar y data flag (ADF)
: data word N
D8
check sum
D8
EP
Value of 0000 0000 to 1111 1111 Sum of D0–D8 of data ID through last data word. Preset to all zeros; carry is ignored.
Notes: EP = even parity for D0–D7.
Table 6.13. EIA-708 Digital Closed Captioning Ancillary Data Packet Format.
Pro-Video Component Interfaces
8-bit Data
115
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
1
1
1
1
0
1
0
0
SDID
D8
EP
0
0
0
0
0
0
0
0
data count (DC)
D8
EP
0
0
0
1
0
0
0
0
D8
EP
crc5
crc4
crc3
crc2
crc1
crc0
0
0
D8
EP
crc11
crc10
crc9
crc8
crc7
crc6
0
0
D8
EP
V
0
crc15
crc14
crc13
crc12
0
0
D8
EP
crc5
crc4
crc3
crc2
crc1
crc0
0
0
D8
EP
crc11
crc10
crc9
crc8
crc7
crc6
0
0
D8
EP
V
0
crc15
crc14
crc13
crc12
0
0
ancillar y flags
D8
EP
0
ues
ida
idh
eda
edh
0
0
active flags
D8
EP
0
ues
ida
idh
eda
edh
0
0
field flags
D8
EP
0
ues
ida
idh
eda
edh
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
ancillar y data flag (ADF)
active CRC
field CRC
reser ved
check sum
D8
Sum of D0–D8 of data ID through last reserved word. Preset to all zeros; carry is ignored.
Notes: EP = even parity for D0–D7.
Table 6.14. Error Detection Ancillary Data Packet Format.
116
Chapter 6: Digital Video Interfaces
The offset value is a 5-bit unsigned integer which represents the offset (in lines) of the data insertion line, relative to line 9 or 272 for 480i systems and line 5 or 318 for 576i systems. The type value identifies the VBI data content, with EIA-516 (DC = 36D; 34 data bytes starting with byte sync value), Guide Plus + (DC = 6D; 4 data bytes), AMOL (DC = 8D; 6 data bytes) and AMOL II (DC = 14D; 12 data bytes) already defined.
VBI Data Ser vice The VBI data service (SMPTE RP-208) is intended for use in reconstructing VBI data in a standard-definition analog video signal produced from the digital video program. Ancillary data packets contain the sampled VBI data, as seen in Table 6.15. The field bit is “0” for Field 2 and “1” for Field 1.
8-bit Data
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
1
0
0
0
0
1
SDID
D8
EP
0
0
0
0
0
0
1
0
data count (DC)
D8
EP
0
0
0
0
0
0
1
1
line
D8
EP
field
0
0
type
D8
EP
type
data word 0
D8
EP
Value of 0000 0000 to 1111 1111
ancillar y data flag (ADF)
offset
: data word N
D8
check sum
D8
EP
Value of 0000 0000 to 1111 1111 Sum of D0–D8 of data ID through last data word. Preset to all zeros; carry is ignored.
Notes: EP = even parity for D0–D7.
Table 6.15. VBI Data Service Ancillary Packet Format.
117
Pro-Video Component Interfaces
Video Index Format If the video index for 480i or 480p systems (SMPTE RP-186) is present, it is carried on the CbCr data channels in the active portion of lines 14 and 277 for 480i systems and lines 27 and 28 for 480p systems. A total of 90 8-bit data words are transferred serially by D2 of the 720 CbCr samples of the active portion of the lines. A 10-bit value of 200H represents a “0;” a 10-bit value of 204H represents a “1.” Unused CbCr samples have a 10-bit value of 200H.
Video Payload Identification Format ITU-R BT.1614 and SMPTE 352M define a 4-byte payload identifier which may be used for identifying the video payload. For 480i digital interfaces, lines 13 and 276 are used to carry the ancillary data packet. For 480p digital interfaces, line 13 is used to carry the ancillary data packet. For 576i digital interfaces, lines 9 and 322 are used to carry the ancillary data packet. For 576p digital interfaces, line 9 is used to carry the ancillary data packet.
8-bit Data
10-bit Data
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
0
0
0
0
0
1
SDID
D8
EP
0
0
0
0
0
0
0
1
data count (DC)
D8
EP
0
0
0
0
0
1
0
0
version ID
ancillar y data flag (ADF)
video payload
D8
EP
picture rate
D8
EP
sampling structure
D8
EP
other info
D8
EP
check sum
D8
payload ID
I/P I/P transport picture 16:9
0
channel
0
0
picture rate
0
0
sampling structure
0
dynamic range
0
Sum of D0–D8 of data ID through other info word. Preset to all zeros; carry is ignored.
Notes: EP = even parity for D0–D7.
Table 6.16. Video Payload ID Ancillary Packet Format.
bit depth
118
Chapter 6: Digital Video Interfaces
For 720p digital interfaces, line 10 is used to carry the ancillary data packet. For 1080i digital interfaces, lines 10 and 572 are used to carry the ancillary data packet. For 1080p digital interfaces, line 10 is used to carry the ancillary data packet. The version ID bit is “1” if based on SMPTE 352M-2002; “0” if based on SMPTE 352M as trial published in the July, 2001 issue of the SMPTE Journal. The I/P transport bit is provided for rapid and reliable detection of the transport scanning structure. The scanning structure can be determined through the “F” bit, which has a static “0” value for progressive transports, and toggles between “0” (field 1) and “1” (field 2) for interlaced transports. However, this detection may take several transport frames to ensure accuracy. The I/P transport bit can provide the information on a per-frame basis. The I/P picture bit is used to identify whether the picture has been scanned as progressive (“1”) or interlaced (“0”). The picture rate bits indicate the picture frame rate in Hz, as specified in Table 6.17. The 16:9 bit is used to identify whether the picture aspect ratio is 4:3 (“0”) or 16:9 (“1”). The sampling structure bits indicate the YCbCr or RGB sampling structure of the picture, as specified in Table 6.18.
The channel bits indicate channel identification information: “00” = single channel, or channel 1 of a multi-channel, video payload “01” = channel 2 of a multi-channel video payload “10” = channel 3 of a multi-channel video payload “11” = channel 4 of a multi-channel video payload
The dynamic range bits identify the dynamic range of the sample quantization: “00” = quantization value range normal “01” = quantization value range extended to 200% of normal (1 overhead bit in the MSB location) “10” = quantization value range extended to 400% of normal (2 overhead bits in the MSB locations) “11” = reserved
The bit depth bits identify the bit depth of the sample quantization: “00” = 8 bits per sample “01” = 10 bits per sample “10” = 12 bits per sample “11” = reserved
Pro-Video Component Interfaces
119
Picture Rate Code (D3-D0)
Picture Rate Value
Picture Rate Code (D3-D0)
Picture Rate Value
Picture Rate Code (D3-D0)
Picture Rate Value
Picture Rate Code (D3-D0)
Picture Rate Value
0000
not defined
0100
reserved
1000
reserved
1100
reserved
0001
reserved
0101
25
1001
50
1101
reserved
0010
24/1.001
0110
30/1.001
1010
60/1.001
1110
reserved
0011
24
0111
30
1011
60
1111
reserved
Table 6.17. Picture Rate Values.
Sampling Structure Code (D3-D0)
Sampling Structure Format
Sampling Structure Code (D3-D0)
Sampling Structure Format
Sampling Structure Code (D3-D0)
Sampling Structure Format
Sampling Structure Code (D3-D0)
Sampling Structure Format
0000
4:2:2 YCbCr
0100
4:2:2:4 YCbCrA
1000
4:2:2:4 YCbCrD
1100
reserved
0001
4:4:4 YCbCr
0101
4:4:4:4 YCbCrA
1001
4:4:4:4 YCbCrD
1101
reserved
0010
4:4:4 GBR
0110
4:4:4:4 GBRA
1010
4:4:4:4 GBRD
1110
reserved
0011
4:2:0 YCbCr
0111
reserved
1011
reserved
1111
reserved
Notes: “A” notation refers to picture channel, “D” notation refers to a non-picture (i.e. data) channel.
Table 6.18. Sampling Structure Formats.
120
Chapter 6: Digital Video Interfaces
Pin
Signal
Pin
Signal
1
clock
14
clock–
2
system ground A
15
system ground B
3
D9
16
D9–
4
D8
17
D8–
5
D7
18
D7–
6
D6
19
D6–
7
D5
20
D5–
8
D4
21
D4–
9
D3
22
D3–
10
D2
23
D2–
11
D1
24
D1–
12
D0
25
D0–
13
cable shield
Table 6.19. 25-Pin Parallel Interface Connector Pin Assignments. For 8-bit interfaces, D9–D2 are used.
25-pin Parallel Interface This interface is used to transfer SDTV resolution 4:2:2 YCbCr data. 8-bit or 10-bit data and a clock are transferred. The individual bits are labeled D0–D9, with D9 being the most significant bit. The pin allocations for the signals are shown in Table 6.19. Y has a nominal 10-bit range of 040H– 3ACH. Values less than 040H or greater than 3ACH may be present due to processing. During blanking, Y data should have a value of 040H, unless other information is present. Cb and Cr have a nominal 10-bit range of 040H–3C0H. Values less than 040H or greater than 3C0H may be present due to processing. During blanking, CbCr data should have a value of 200H, unless other data is present. Signal levels are compatible with ECLcompatible balanced drivers and receivers. The generator must have a balanced output
with a maximum source impedance of 110 Ω; the signal must be 0.8–2.0V peak-to-peak measured across a 110-Ω load. At the receiver, the transmission line is terminated by 110 ±10 Ω. 27 MHz Parallel Interface This BT.656 and SMPTE 125M interface is used for 480i and 576i systems with an aspect ratio of 4:3. Y and multiplexed CbCr information at a sample rate of 13.5 MHz are multiplexed into a single 8-bit or 10-bit data stream, at a clock rate of 27 MHz. The 27 MHz clock signal has a clock pulse width of 18.5 ±3 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±3 ns (as shown in Figure 6.2). To permit reliable operation at interconnect lengths of 50–200 meters, the receiver must use frequency equalization, with typical
Pro-Video Component Interfaces
CLOCK TW TC
TD
DATA
TW = 18.5 ± 3 NS TC = 37 NS TD = 18.5 ± 3 NS
Figure 6.2. 25-Pin 27 MHz Parallel Interface Waveforms. RELATIVE GAIN (DB) 20
18
16
14
12
10
8
6
4
2
0 0.1
1
10
100
FREQUENCY (MHZ)
Figure 6.3. Example Line Receiver Equalization Characteristics for Small Signals.
121
122
Chapter 6: Digital Video Interfaces
characteristics shown in Figure 6.3. This example enables operation with a range of cable lengths down to zero. 36 MHz Parallel Interface This BT.1302 and SMPTE 267M interface is used for 480i and 576i systems with an aspect ratio of 16:9. Y and multiplexed CbCr information at a sample rate of 18 MHz are multiplexed into a single 8-bit or 10-bit data stream, at a clock rate of 36 MHz. The 36 MHz clock signal has a clock pulse width of 13.9 ±2 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±2 ns (as shown in Figure 6.4. To permit reliable operation at interconnect lengths of 40–160 meters, the receiver must use frequency equalization, with typical characteristics shown in Figure 6.3.
93-pin Parallel Interface This interface is used to transfer 16:9 HDTV resolution R´G´B´ data, 4:2:2 YCbCr data, or 4:2:2:4 YCbCrK data. The pin allocations for the signals are shown in Table 6.20. The most significant bits are R9, G9, and B9. When transferring 4:2:2 YCbCr data, the green channel carries Y information and the red channel carries multiplexed CbCr information. When transferring 4:2:2:4 YCbCrK data, the green channel carries Y information, the red channel carries multiplexed CbCr information, and the blue channel carries K (alpha keying) information. Y has a nominal 10-bit range of 040H– 3ACH. Values less than 040H or greater than 3ACH may be present due to processing. During blanking, Y data should have a value of 040H, unless other information is present.
CLOCK TW TD
TC
DATA
TW = 13.9 ± 2 NS TC = 27.8 NS TD = 13.9 ± 2 NS
Figure 6.4. 25-Pin 36 MHz Parallel Interface Waveforms.
Pro-Video Component Interfaces
Pin
Signal
Pin
Signal
Pin
Signal
Pin
Signal
1
clock
26
GND
51
B2
76
GND
2
G9
27
GND
52
B1
77
GND
3
G8
28
GND
53
B0
78
GND
4
G7
29
GND
54
R9
79
B4–
5
G6
30
GND
55
R8
80
B3–
6
G5
31
GND
56
R7
81
B2–
7
G4
32
GND
57
R6
82
B1–
8
G3
33
clock–
58
R5
83
B0–
9
G2
34
G9–
59
R4
84
R9–
10
G1
35
G8–
60
R3
85
R8–
11
G0
36
G7–
61
R2
86
R7–
12
B9
37
G6–
62
R1
87
R6–
13
B8
38
G5–
63
R0
88
R5–
14
B7
39
G4–
64
GND
89
R4–
15
B6
40
G3–
65
GND
90
R3–
16
B5
41
G2–
66
GND
91
R2–
17
GND
42
G1–
67
GND
92
R1–
93
R0–
18
GND
43
G0–
68
GND
19
GND
44
B9–
69
GND
20
GND
45
B8–
70
GND
21
GND
46
B7–
71
GND
22
GND
47
B6–
72
GND
23
GND
48
B5–
73
GND
24
GND
49
B4
74
GND
25
GND
50
B3
75
GND
Table 6.20. 93-Pin Parallel Interface Connector Pin Assignments. For 8-bit interfaces, bits 9–2 are used.
123
124
Chapter 6: Digital Video Interfaces
CLOCK TW TD
TC
DATA
TW = 6.73 ± 1.48 NS TC = 13.47 NS TD = 6.73 ± 1 NS
Figure 6.5. 93-Pin 74.25 MHz Parallel Interface Waveforms.
Cb and Cr have a nominal 10-bit range of 040H–3C0H. Values less than 040H or greater than 3C0H may be present due to processing. During blanking, CbCr data should have a value of 200H, unless other information is present. R´G´B´ and K have a nominal 10-bit range of 040H–3ACH. Values less than 040H or greater than 3ACH may be present due to processing. During blanking, R´G´B´ data should have a value of 040H, unless other information is present. Signal levels are compatible with ECLcompatible balanced drivers and receivers. The generator must have a balanced output with a maximum source impedance of 110 Ω; the signal must be 0.6–2.0V peak-to-peak measured across a 110-Ω load. At the receiver, the transmission line must be terminated by 110 ±10 Ω.
74.25 MHz Parallel Interface This ITU-R BT.1120 and SMPTE 274M interface is primarily used for 16:9 HDTV systems. The 74.25 MHz clock signal has a clock pulse width of 6.73 ±1.48 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±1 ns (as shown in Figure 6.5). To permit reliable operation at interconnect lengths greater than 20 meters, the receiver must use frequency equalization. 74.176 MHz Parallel Interface This BT.1120 and SMPTE 274M interface is primarily used for 16:9 HDTV systems. The 74.176 MHz (74.25/1.001) clock signal has a clock pulse width of 6.74 ±1.48 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±1 ns (similar to Figure 6.5). To permit reliable operation at interconnect lengths greater than 20 meters, the receiver must use frequency equalization.
Pro-Video Component Interfaces
148.5 MHz Parallel Interface This BT.1120 and SMPTE 274M interface is used for 16:9 HDTV systems. The 148.5 MHz clock signal has a clock pulse width of 3.37 ±0.74 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±0.5 ns (similar to Figure 6.5). To permit reliable operation at interconnect lengths greater than 14 meters, the receiver must use frequency equalization. 148.35 MHz Parallel Interface This BT.1120 and SMPTE 274M interface is used for 16:9 HDTV systems. The 148.35 MHz (148.5/1.001) clock signal has a clock pulse width of 3.37 ±0.74 ns. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±0.5 ns (similar to Figure 6.5). To permit reliable operation at interconnect lengths greater than 14 meters, the receiver must use frequency equalization.
Serial Interfaces The parallel formats can be converted to a serial format (Figure 6.6), allowing data to be transmitted using a 75-Ω coaxial cable (or optical fiber). Equipment inputs and outputs both use BNC connectors so that interconnect cables can be used in either direction. For cable interconnect, the generator has an unbalanced output with a source impedance of 75Ω; the signal must be 0.8V ±10% peak-topeak measured across a 75-Ω load. The receiver has an input impedance of 75Ω. In an 8-bit environment, before serialization, the 00H and FFH codes during EAV and
125
SAV are expanded to 10-bit values of 000H and 3FFH, respectively. All other 8-bit data is appended with two least significant “0” bits before serialization. The 10 bits of data are serialized (LSB first) and processed using a scrambled and polarity-free NRZI algorithm: G(x) = (x9 + x4 + 1)(x + 1)
The input signal to the scrambler (Figure 6.7) uses positive logic (the highest voltage represents a logical one; lowest voltage represents a logical zero). The formatted serial data is output at the 10× sample clock rate. Since the parallel clock may contain large amounts of jitter, deriving the 10× sample clock directly from an unfiltered parallel clock may result in excessive signal jitter. At the receiver, phase-lock synchronization is done by detecting the EAV and SAV sequences. The PLL is continuously adjusted slightly each scan line to ensure that these patterns are detected and to avoid bit slippage. The recovered 10× sample clock is divided by ten to generate the sample clock, although care must be taken not to mask word-related jitter components. The serial data is low- and high-frequency equalized, inverse scrambling performed (Figure 6.8), and deserialized. 270 Mbps Serial Interface This BT.656 and SMPTE 259M interface (also called SDI) converts a 27 MHz parallel stream into a 270 Mbps serial stream. The 10× PLL generates a 270 MHz clock from the 27 MHz clock signal. This interface is primarily used for 480i and 576i 4:3 systems.
126
PARALLEL 4:2:2 VIDEO
Chapter 6: Digital Video Interfaces
75–OHM COAX
10
SHIFT REGISTER
SCRAMBLER
10
SHIFT REGISTER
DESCRAMBLER
SAMPLE CLOCK
PARALLEL 4:2:2 VIDEO
SAV, EAV DETECT SERIAL
10X PLL
CLOCK SAMPLE CLOCK
DIVIDE BY 10
PLL
Figure 6.6. Serial Interface Block Diagram.
SERIAL DATA IN (NRZ)
+
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
+
Q
D
Q
ENCODED DATA OUT (NRZI)
+ G(X) = X9 + X4 + 1
G(X) = X + 1
Figure 6.7. Typical Scrambler Circuit.
ENCODED DATA IN (NRZI)
D
Q
+
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
+
Figure 6.8. Typical Descrambler Circuit.
D
Q
D
Q
+
SERIAL DATA OUT (NRZ)
127
Pro-Video Component Interfaces
360 Mbps Serial Interface This BT.1302 and SMPTE 259M interface converts a 36 MHz parallel stream into a 360 Mbps serial stream. The 10× PLL generates a 360 MHz clock from the 36 MHz clock signal. This interface is primarily used for 480i and 576i 16:9 systems. 540 Mbps Serial Interface This SMPTE 344M interface converts a 54 MHz parallel stream, or two 27 MHz parallel streams, into a 540 Mbps serial stream. The 10× PLL generates a 540 MHz clock from the 54 MHz clock signal. This interface is primarily used for 480p and 576p 4:3 systems. 1.485 Gbps Serial Interface This BT.1120 and SMPTE 292M interface multiplexes two 74.25 MHz parallel streams (Y and CbCr) into a single 1.485 Gbps serial stream. A 20× PLL generates a 1.485 GHz clock from the 74.25 MHz clock signal. This interface is used for 16:9 HDTV systems.
D9 (MSB)
D8
D7
D6
LN0
D8
L6
L5
LN1
D8
0
0
Before multiplexing the two parallel streams together, line number and CRC information (Table 6.21) is added to each stream after each EAV sequence. The CRC is used to detect errors in the active video and EAV. It consists of two words generated by the polynomial: CRC = x18 + x5 + x4 + 1
The initial value is set to zero. The calculation starts with the first active line word and ends at the last word of the line number (LN1). 1.4835 Gbps Serial Interface This BT.1120 and SMPTE 292M interface multiplexes two 74.176 (74.25/1.001) MHz parallel streams (Y and CbCr) into a single 1.4835 (1.485/1.001) Gbps serial stream. A 20× PLL generates a 1.4835 GHz clock from the 74.176 MHz clock signal. This interface is used for 16:9 HDTV systems. Line number and CRC information is added as described for the 1.485 Gbps serial interface.
D5
D4
D3
D2
D1
D0
L4
L3
L2
L1
L0
0
0
0
L10
L9
L8
L7
0
0
CRC0
D8
crc8
crc7
crc6
crc5
crc4
crc3
crc2
crc1
crc0
CRC1
D8
crc17
crc16
crc15
crc14
crc13
crc12
crc11
crc10
crc9
Table 6.21. Line Number and CRC Data.
128
Chapter 6: Digital Video Interfaces
SDTV—Interlaced Supported active resolutions, with their corresponding aspect ratios and frame refresh rates, are: 720 × 480i 720 × 576i 960 × 480i 960 × 576i
4:3 4:3 16:9 16:9
29.97 Hz 25.00 Hz 29.97 Hz 25.00 Hz
4:2:2 YCbCr Parallel Interface The ITU-R BT.656 and BT.1302 parallel interfaces were developed to transfer BT.601 4:2:2 YCbCr digital video between equipment. SMPTE 125M and 267M further clarify the operation for 480i systems. Figure 6.9 illustrates the timing for one scan line for the 4:3 aspect ratio, using a 27 MHz sample clock. Figure 6.10 shows the timing for one scan line for the 16:9 aspect ratio, using a 36 MHz sample clock. The 25-pin parallel interface is used. 4:2:2 YCbCr Serial Interface BT.656 and BT.1302 also define a YCbCr serial interface. The 10-bit 4:2:2 YCbCr parallel streams shown in Figures 6.9 and 6.10 are serialized using the 270 or 360 Mbps serial interface. 4:4:4:4 YCbCrK Parallel Interface The ITU-R BT.799 and BT.1303 parallel interfaces were developed to transfer BT.601 4:4:4:4 YCbCrK digital video between equipment. K is an alpha keying signal, used to mix two video sources, discussed in Chapter 7. SMPTE RP-175 further clarifies the operation for 480i systems.
Multiplexing Structure Two transmission links are used. Link A contains all the Y samples plus those Cb and Cr samples located at even-numbered sample points. Link B contains samples from the keying channel and the Cb and Cr samples from the odd-numbered sampled points. Although it may be common to refer to Link A as 4:2:2 and Link B as 2:2:4, Link A is not a true 4:2:2 signal since the CbCr data was sampled at 13.5 MHz, rather than 6.75 MHz. Figure 6.11 shows the contents of links A and B when transmitting 4:4:4:4 YCbCrK video data. Figure 6.12 illustrates the contents when transmitting R´G´B´K video data. If the keying signal (K) is not present, the K sample values should have a 10-bit value of 3ACH. Figure 6.13 illustrates the YCbCrK timing for one scan line for the 4:3 aspect ratio, using a 27 MHz sample clock. Figure 6.14 shows the YCbCrK timing for one scan line for the 16:9 aspect ratio, using a 36 MHz sample clock. Two 25-pin parallel interfaces are used. 4:4:4:4 YCbCrK Serial Interface BT.799 and BT.1303 also define a YCbCr serial interface. The two 10-bit 4:2:2 YCbCr parallel streams shown in Figure 6.13 or 6.14 are serialized using two 270 or 360 Mbps serial interfaces. SMPTE RP-175 further clarifies the operation for 480i systems. RGBK Parallel Interface BT.799 and BT.1303 also support transferring BT.601 R´G´B´K digital video between equipment. For additional information, see the 4:4:4:4 YCbCrK parallel interface. SMPTE RP175 further clarifies the operation for 480i systems. The G´ samples are sent in the Y locations, the R´ samples are sent in the Cr locations, and the B´ samples are sent in the Cb locations.
129
Pro-Video Component Interfaces
BT.601 H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
4
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
0 0 0
268 (280)
0 0 0
CO–SITED X Y Z
C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
4
C R 2
Y 3
C R 718
Y 719
3 F F
BT.656 4:2:2 VIDEO
1440
1716 (1728)
Figure 6.9. BT.656 and SMPTE 125M Parallel Interface Data For One Scan Line. 480i; 4:2:2 YCbCr; 720 active samples per line; 27 MHz clock; 10-bit system. The values for 576i systems are shown in parentheses.
BT.601 H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0 4
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
360 (376)
0 0 0
0 0 0 4
CO–SITED X Y Z
C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
C R 2
Y 3
C R 958
Y 959
3 F F
1920
2288 (2304)
Figure 6.10. BT.1302 and SMPTE 267M Parallel Interface Data For One Scan Line. 480i; 4:2:2 YCbCr; 960 active samples per line; 36 MHz clock; 10-bit system. The values for 576i systems are shown in parentheses.
BT.1302 4:2:2 VIDEO
130
Chapter 6: Digital Video Interfaces
SAMPLE NUMBER
LINK A
SAMPLE NUMBER
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
Y
Y
Y
Y
Y
Y
Y
Y
G
G
G
G
G
G
G
G
CB
CB
CB
CB
CB
CB
CB
CB
B
B
B
B
B
B
B
B
CR
CR
CR
CR
CR
CR
CR
CR
R
R
R
R
R
R
R
R
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
LINK B
LINK A
LINK B
Figure 6.11. Link Content Representation for YCbCrK Video Signals.
Figure 6.12. Link Content Representation for R´G´B´K Video Signals.
BT.601 H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
4
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
0 0 0
268 (280)
0 0 0
CO–SITED X Y Z
C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
4
C R 2
Y 3
C R 718
Y 719
3 F F
4:2:2 STREAM (LINK A)
K 3
C R 719
K 719
3 F F
4:2:2 STREAM (LINK B)
1440
1716 (1728)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
0 0 0
0 0 0
X Y Z
C B 1
K 0
C R 1
K 1
C B 3
K 2
C R 3
Figure 6.13. BT.799 and SMPTE RP-175 Parallel Interface Data For One Scan Line. 480i; 4:4:4:4 YCbCrK; 720 active samples per line; 27 MHz clock; 10-bit system. The values for 576i systems are shown in parentheses.
Pro-Video Component Interfaces
131
BT.601 H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
4
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
0 0 0
360 (376)
0 0 0
CO–SITED X Y Z
C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
4
C R 2
Y 3
C R 958
Y 959
3 F F
4:2:2 STREAM (LINK A)
K 3
C R 959
K 959
3 F F
4:2:2 STREAM (LINK B)
1920
2288 (2304)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
0 4 0
SAV CODE 2 0 0
0 4 0
3 F F
0 0 0
0 0 0
X Y Z
C B 1
K 0
C R 1
K 1
C B 3
K 2
C R 3
Figure 6.14. BT.1303 Parallel Interface Data For One Scan Line. 480i; 4:4:4:4 YCbCrK; 960 active samples per line; 36 MHz clock; 10-bit system. The values for 576i systems are shown in parentheses.
RGBK Serial Interface BT.799 and BT.1303 also define a R´G´B´K serial interface. The two 10-bit R´G´B´K parallel streams are serialized using two 270 or 360 Mbps serial interfaces.
SDTV—Progressive Supported active resolutions, with their corresponding aspect ratios and frame refresh rates, are: 720 × 480p 720 × 576p
4:3 4:3
59.94 Hz 50.00 Hz
4:2:2 YCbCr Serial Interface ITU-R BT.1362 defines two 10-bit 4:2:2 YCbCr data streams (Figure 6.15), using a 27 MHz sample clock. SMPTE 294M further clarifies the operation for 480p systems. What stream is used for which scan line is shown in Table 6.22. The two 10-bit parallel streams shown in Figure 6.15 are serialized using two 270 Mbps serial interfaces.
132
Chapter 6: Digital Video Interfaces
BT.1358, SMPTE 293M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
0 4 0
2 0 0
4
SAV CODE
268 (280)
X Y Z
0 0 0
0 0 0
3 F F
0 4 0
CO–SITED C B 0
Y 0
C R 0
NEXT LINE
CO–SITED Y 1
C B 2
Y 2
4
C R 2
Y 3
C R 718
Y 719
3 F F
BT.656 4:2:2 STREAM (LINK A)
Y 3
C R 718
Y 719
3 F F
BT.656 4:2:2 STREAM (LINK B)
1440
1716 (1728)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
0 4 0
2 0 0
SAV CODE
0 4 0
2 0 0
0 4 0
3 F F
0 0 0
0 0 0
X Y Z
C B 0
Y 0
C R 0
Y 1
C B 2
Y 2
C R 2
Figure 6.15. BT.1362 and SMPTE 294M Parallel Data For Two Scan Lines. 480p; 4:2:2 YCbCr; 720 active samples per line; 27 MHz clock; 10-bit system. The values for 576p systems are shown in parentheses.
480p (525p) System
576p (625p) System
Link A
Link B
Link A
Link B
Link A
Link B
Link A
Link B
7
8
6
7
1
2
4
5
9
10
:
:
3
4
6
7
:
:
522
523
:
:
8
9
523
524
524
525
621
622
:
:
525
1
1
2
623
624
620
621
2
3
3
4
625
1
622
623
4
5
5
6
2
3
624
625
Table 6.22. BT.1362 and SMPTE 294M Scan Line Numbering and Link Assignment.
Pro-Video Component Interfaces
4:2:2 YCbCr Serial Interface BT.1120 also defines a YCbCr serial interface. SMPTE 292M further clarifies the operation for 29.97 and 30 Hz systems. The two 10bit 4:2:2 YCbCr parallel streams shown in Figure 6.16 are multiplexed together, then serialized using a 1.485 or 1.4835 Gbps serial interface.
HDTV—Interlaced Supported active resolutions, with their corresponding aspect ratios and frame refresh rates, are: 1920 × 1080i 1920 × 1080i 1920 × 1080i
16:9 16:9 16:9
133
25.00 Hz 29.97 Hz 30.00 Hz
4:2:2:4 YCbCrK Parallel Interface BT.1120 also supports transferring HDTV 4:2:2:4 YCbCrK digital video between equipment. SMPTE 274M further clarifies the operation for 29.97 and 30 Hz systems. Figure 6.17 illustrates the timing for one scan line for the 1920 × 1080i active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (25 or 30 Hz refresh) or 74.176 MHz (29.97 Hz refresh).
4:2:2 YCbCr Parallel Interface The ITU-R BT.1120 parallel interface was developed to transfer interlaced HDTV 4:2:2 YCbCr digital video between equipment. SMPTE 274M further clarifies the operation for 29.97 and 30 Hz systems. Figure 6.16 illustrates the timing for one scan line for the 1920 × 1080i active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (25 or 30 Hz refresh) or 74.176 MHz (29.97 Hz refresh).
BT.709, SMPTE 274M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
272 (712)
0 0 0
X Y Z
Y 0
Y 1
Y 2
Y 3
Y 4
Y 5
4
Y 6
Y 7
Y 1918
Y 1919
3 F F
Y CHANNEL
C R 6
C B 1918
C R 1918
3 F F
CBCR CHANNEL
1920
2200 (2640)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
2 0 0
2 0 0
2 0 0
SAV CODE 2 0 0
2 0 0
3 F F
0 0 0
0 0 0
X Y Z
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
Figure 6.16. BT.1120 and SMPTE 274M Parallel Interface Data For One Scan Line. 1080i; 29.97-, 30-, 59.94-, and 60-Hz systems; 4:2:2 YCbCr; 1920 active samples per line; 74.176, 74.25, 148.35, or 148.5 MHz clock; 10-bit system. The values for 25and 50-Hz systems are shown in parentheses.
134
Chapter 6: Digital Video Interfaces
BT.709, SMPTE 274M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
272 (712)
0 0 0
X Y Z
Y 0
Y 1
Y 2
Y 3
Y 4
Y 5
4
Y 6
Y 7
Y 1918
Y 1919
3 F F
Y CHANNEL
1920
2200 (2640)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
2 0 0
2 0 0
EAV CODE 3 F F
0 0 0
0 0 0
2 0 0
SAV CODE 2 0 0
2 0 0
3 F F
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
0 4 0
0 0 0
0 0 0
X Y Z
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
C R 6
C B 1918
C R 1918
3 F F
CBCR CHANNEL
X Y Z
K 0
K 1
K 2
K 3
K 4
K 5
K 6
K 7
K 1918
K 1919
3 F F
K CHANNEL
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
0 0 0
Figure 6.17. BT.1120 and SMPTE 274M Parallel Interface Data For One Scan Line. 1080i; 29.97-, 30-, 59.94-, and 60-Hz systems; 4:2:2:4 YCbCrK; 1920 active samples per line; 74.176, 74.25, 148.35, or 148.5 MHz clock; 10-bit system. The values for 25and 50-Hz systems are shown in parentheses.
Pro-Video Component Interfaces
135
Figure 6.18 illustrates the timing for one scan line for the 1920 × 1080i active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (25 or 30 Hz refresh) or 74.176 MHz (29.97 Hz refresh).
RGB Parallel Interface BT.1120 also supports transferring HDTV R´G´B´ digital video between equipment. SMPTE 274M further clarifies the operation for 29.97 and 30 Hz systems.
BT.709, SMPTE 274M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
272 (712)
0 0 0
X Y Z
G 0
G 1
G 2
G 3
G 4
G 5
4
G 6
G 7
G 1918
G 1919
3 F F
GREEN CHANNEL
1920
2200 (2640)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
X Y Z
0 4 0
0 4 0
0 4 0
EAV CODE 3 F F
0 0 0
0 0 0
0 4 0
SAV CODE 0 4 0
0 4 0
3 F F
0 4 0
0 4 0
3 F F
BLANKING 0 4 0
0 0 0
0 0 0
X Y Z
R 0
R 1
R 2
R 3
R 4
R 5
R 6
R 7
R 1918
R 1919
3 F F
RED CHANNEL
X Y Z
B 0
B 1
B 2
B 3
B 4
B 5
B 6
B 7
B 1918
B 1919
3 F F
BLUE CHANNEL
SAV CODE 0 0 0
0 0 0
Figure 6.18. BT.1120 and SMPTE 274M Parallel Interface Data For One Scan Line. 1080i; 29.97-, 30-, 59.94-, and 60-Hz systems; R´G´B´; 1920 active samples per line; 74.176, 74.25, 148.35, or 148.5 MHz clock; 10-bit system. The values for 25- and 50Hz systems are shown in parentheses.
136
Chapter 6: Digital Video Interfaces
HDTV—Progressive Supported active resolutions, with their corresponding aspect ratios and frame refresh rates, are: 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p 1280 × 720p
1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p 1920 × 1080p
16:9 16:9 16:9 16:9 16:9 16:9 16:9 16:9
23.98 Hz 24.00 Hz 25.00 Hz 29.97 Hz 30.00 Hz 50.00 Hz 59.94 Hz 60.00 Hz
16:9 16:9 16:9 16:9 16:9 16:9 16:9 16:9
23.98 Hz 24.00 Hz 25.00 Hz 29.97 Hz 30.00 Hz 50.00 Hz 59.94 Hz 60.00 Hz
4:2:2 YCbCr Parallel Interface The ITU-R BT.1120 and SMPTE 274M parallel interfaces were developed to transfer progressive HDTV 4:2:2 YCbCr digital video between equipment. Figure 6.16 illustrates the timing for one scan line for the 1920 × 1080p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 148.5 MHz (24, 25, 30, 50 or 60 Hz refresh) or 148.35 MHz (23.98, 29.97 or 59.94 Hz refresh). Figure 6.19 illustrates the timing for one scan line for the 1280 × 720p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (24, 25, 30, 50
or 60 Hz refresh) or 74.176 MHz (23.98, 29.97, or 59.94 Hz refresh). 4:2:2:4 YCbCrK Parallel Interface BT.1120 and SMPTE 274M also support transferring HDTV 4:2:2:4 YCbCrK digital video between equipment. Figure 6.17 illustrates the timing for one scan line for the 1920 × 1080p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 148.5 MHz (24, 25, 30, 50 or 60 Hz refresh) or 148.35 MHz (23.98, 29.97 or 59.94 Hz refresh). Figure 6.20 illustrates the timing for one scan line for the 1280 × 720p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (24, 25, 30, 50 or 60 Hz refresh) or 74.176 MHz (23.98, 29.97 or 59.94 Hz refresh). RGB Parallel Interface BT.1120 and SMPTE 274M also support transferring HDTV R´G´B´ digital video between equipment. Figure 6.18 illustrates the timing for one scan line for the 1920 × 1080p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 148.5 MHz (24, 25, 30, 50 or 60 Hz refresh) or 148.35 MHz (23.98, 29.97 or 59.94 Hz refresh). Figure 6.21 illustrates the timing for one scan line for the 1280 × 720p active resolutions. The 93-pin parallel interface is used with a sample clock rate of 74.25 MHz (24, 25, 30, 50 or 60 Hz refresh) or 74.176 MHz (23.98, 29.97 or 59.94 Hz refresh).
Pro-Video Component Interfaces
137
SMPTE 296M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
362 (692)
0 0 0
X Y Z
Y 0
Y 1
Y 2
Y 3
Y 4
Y 5
4
Y 6
Y 7
Y 1278
Y 1279
3 F F
Y CHANNEL
C R 6
C B 1278
C R 1278
3 F F
CBCR CHANNEL
1280
1650 (1980)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
2 0 0
2 0 0
2 0 0
SAV CODE 2 0 0
2 0 0
3 F F
0 0 0
0 0 0
X Y Z
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
Figure 6.19. SMPTE 274M Parallel Interface Data For One Scan Line. 720p; 59.94and 60-Hz systems; 4:2:2 YCbCr; 1280 active samples per line; 74.176 or 74.25 MHz clock; 10-bit system. The values for 50-Hz systems are shown in parentheses. SMPTE 296M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
362 (692)
0 0 0
X Y Z
Y 0
Y 1
Y 2
Y 3
Y 4
Y 5
4
Y 6
Y 7
Y 1278
Y 1279
3 F F
Y CHANNEL
1280
1650 (1980)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
2 0 0
2 0 0
2 0 0
X Y Z
0 4 0
0 4 0
0 4 0
EAV CODE 3 F F
0 0 0
0 0 0
2 0 0
SAV CODE 2 0 0
2 0 0
3 F F
0 4 0
0 4 0
3 F F
BLANKING 0 4 0
0 0 0
0 0 0
X Y Z
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
C R 6
C B 1278
C R 1278
3 F F
CBCR CHANNEL
X Y Z
K 0
K 1
K 2
K 3
K 4
K 5
K 6
K 7
K 1278
K 1279
3 F F
K CHANNEL
SAV CODE 0 0 0
0 0 0
Figure 6.20. SMPTE 274M Parallel Interface Data For One Scan Line. 720p; 59.94and 60-Hz systems; 4:2:2:4 YCbCrK; 1280 active samples per line; 74.176 or 74.25 MHz clock; 10-bit system. The values for 50-Hz systems are shown in parentheses.
138
Chapter 6: Digital Video Interfaces
SMPTE 296M H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
4
0 4 0
NEXT LINE
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
362 (692)
0 0 0
X Y Z
G 0
G 1
G 2
G 3
G 4
G 5
4
G 6
G 7
G 1278
G 1279
3 F F
GREEN CHANNEL
1280
1650 (1980)
EAV CODE 3 F F
0 0 0
0 0 0
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
EAV CODE 3 F F
0 0 0
0 0 0
0 4 0
SAV CODE 0 4 0
0 4 0
3 F F
BLANKING X Y Z
0 4 0
0 4 0
0 4 0
0 4 0
0 0 0
0 0 0
X Y Z
R 0
R 1
R 2
R 3
R 4
R 5
R 6
R 7
R 1278
R 1279
3 F F
RED CHANNEL
X Y Z
B 0
B 1
B 2
B 3
B 4
B 5
B 6
B 7
B 1278
B 1279
3 F F
BLUE CHANNEL
SAV CODE 0 4 0
0 4 0
3 F F
0 0 0
0 0 0
Figure 6.21. SMPTE 274M Parallel Interface Data For One Scan Line. 720p; 59.94and 60-Hz systems; R´G´B´; 1280 active samples per line; 74.176 or 74.25 MHz clock; 10-bit system. The values for 50-Hz systems are shown in parentheses.
Pro-Video Composite Interfaces
Pro-Video Composite Interfaces
139
NTSC Video Timing There are 910 total samples per scan line, as shown in Figure 6.22. Horizontal count 0 corresponds to the start of active video, and a horizontal count of 768 corresponds to the start of horizontal blanking. Sampling is along the ±I and ±Q axes (33°, 123°, 213°, and 303°). The sampling phase at horizontal count 0 of line 10, Field 1 is on the +I axis (123°). The sync edge values, and the horizontal counts at which they occur, are defined as shown in Figure 6.23 and Tables 6.24–6.26. 8bit values for one color burst cycle are 45, 83, 75, and 37. The burst envelope starts at horizontal count 857, and lasts for 43 clock cycles, as shown in Table 6.24. Note that the peak amplitudes of the burst are not sampled.
Digital composite video is essentially a digital version of a composite analog (M) NTSC or (B, D, G, H, I) PAL video signal. The sample clock rate is four times FSC: about 14.32 MHz for (M) NTSC and about 17.73 MHz for (B, D, G, H, I) PAL. Usually, both 8-bit and 10-bit interfaces are supported, with the 10-bit interface used to transmit 2 bits of fractional video data to minimize cumulative processing errors and to support 10-bit ancillary data. Table 6.23 lists the digital composite levels. Video data may not use the 10-bit values of 000H–003H and 3FCH–3FFH, or the 8-bit values of 00H and FFH, since they are used for timing information.
Video Level
(M) NTSC
(B, D, G, H, I) PAL
peak chroma
972
1040 (limited to 1023)
white
800
844
peak burst
352
380
black
280
256
blank
240
256
peak burst
128
128
peak chroma
104
128
sync
16
4
Table 6.23. 10-Bit Video Levels for Digital Composite Video Signals.
140
Chapter 6: Digital Video Interfaces
DIGITAL BLANKING
DIGITAL ACTIVE LINE
142 SAMPLES (768–909)
768 SAMPLES (0–767)
TOTAL LINE 910 SAMPLES (0–909)
Figure 6.22. Digital Composite (M) NTSC Analog and Digital Timing Relationship.
END OF ANALOG LINE
END OF DIGITAL LINE
768 (60) 784 (41) 50% 785 (17)
787 (4)
Figure 6.23. Digital Composite (M) NTSC Sync Timing. The horizontal counts are shown with the corresponding 8-bit sample values in parentheses.
Pro-Video Composite Interfaces
8-bit Hex Value
Sample 768–782 783 784 785 786 787–849 850 851 852 853 854–856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873
10-bit Hex Value
Fields 1, 3
Fields 2, 4
Fields 1, 3
Fields 2, 4
3C 3A 29 11 04 04 06 17 2F 3C 3C 3C 3D 37 36 4B 49 25 2D 53 4B 25 2D 53 4B 25 2D 53
3C 3A 29 11 04 04 06 17 2F 3C 3C 3C 3B 41 42 2D 2F 53 4B 25 2D 53 4B 25 2D 53 4B 25
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0F0 0F4 0DC 0D6 12C 123 096 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0F0 0EC 104 10A 0B4 0BD 14A 12D 092 0B3 14E 12D 092 0B3 14E 12D 092
Table 6.24a. Digital Values During the Horizontal Blanking Intervals for Digital Composite (M) NTSC Video Signals.
141
142
Chapter 6: Digital Video Interfaces
8-bit Hex Value
Sample 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900–909
10-bit Hex Value
Fields 1, 3
Fields 2, 4
Fields 1, 3
Fields 2, 4
4B 25 2D 53 4B 25 2D 53 4B 25 2D 53 4B 25 2D 53 4B 25 2D 53 4A 2A 33 44 3F 3B 3C
2D 53 4B 25 2D 53 4B 25 2D 53 4B 25 2D 53 4B 25 2D 53 4B 25 2E 4E 45 34 39 3D 3C
12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 129 0A6 0CD 112 0FA 0EC 0F0
0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B3 14E 12D 092 0B7 13A 113 0CE 0E6 0F4 0F0
Table 6.24b. Digital Values During the Horizontal Blanking Intervals for Digital Composite (M) NTSC Video Signals.
Pro-Video Composite Interfaces
Fields 1, 3
Fields 2, 4
Sample
8-bit Hex Value
10-bit Hex Value
Sample
8-bit Hex Value
10-bit Hex Value
768–782 783 784 785 786 787–815 816 817 818 819 820–327 328 329 330 331 332–360 361 362 363 364 365–782
3C 3A 29 11 04 04 06 17 2F 3C 3C 3A 29 11 04 04 06 17 2F 3C 3C
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0
313–327 328 329 330 331 332–360 361 362 363 364 365–782 783 784 785 786 787–815 816 817 818 819 820–327
3C 3A 29 11 04 04 06 17 2F 3C 3C 3A 29 11 04 04 06 17 2F 3C 3C
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0
Table 6.25. Equalizing Pulse Values During the Vertical Blanking Intervals for Digital Composite (M) NTSC Video Signals.
143
144
Chapter 6: Digital Video Interfaces
Fields 1, 3
Fields 2, 4
Sample
8-bit Hex Value
10-bit Hex Value
Sample
8-bit Hex Value
10-bit Hex Value
782 783 784 785 786 787–260 261 262 263 264 265–327 328 329 330 331 332–715 716 717 718 719 720–782
3C 3A 29 11 04 04 06 17 2F 3C 3C 3A 29 11 04 04 06 17 2F 3C 3C
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0
327 328 329 330 331 332–715 716 717 718 719 720–782 783 784 785 786 787–260 261 262 263 264 265–327
3C 3A 29 11 04 04 06 17 2F 3C 3C 3A 29 11 04 04 06 17 2F 3C 3C
0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0 0E9 0A4 044 011 010 017 05C 0BC 0EF 0F0
Table 6.26. Serration Pulse Values During the Vertical Blanking Intervals for Digital Composite (M) NTSC Video Signals.
Pro-Video Composite Interfaces
To maintain zero SCH phase, horizontal count 784 occurs 25.6 ns (33° of the subcarrier phase) before the 50% point of the falling edge of horizontal sync, and horizontal count 785 occurs 44.2 ns (57° of the subcarrier phase) after the 50% point of the falling edge of horizontal sync.
PAL Video Timing There are 1135 total samples per line, except for two lines per frame which have 1137 samples per line, making a total of 709,379 samples per frame. Figure 6.24 illustrates the typical line timing. Horizontal count 0 corresponds to the start of active video, and a horizontal count of 948 corresponds to the start of horizontal blanking.
DIGITAL BLANKING 187 SAMPLES (948–1134)
145
Sampling is along the ±U and ±V axes (0°, 90°, 180°, and 270°), with the sampling phase at horizontal count 0 of line 1, Field 1 on the +V axis (90°). 8-bit color burst values are 95, 64, 32, and 64, continuously repeated. The swinging burst causes the peak burst (32 and 95) and zero burst (64) samples to change places. The burst envelope starts at horizontal count 1058, and lasts for 40 clock cycles. Sampling is not H-coherent as with (M) NTSC, so the position of the sync pulses change from line to line. Zero SCH phase is defined when alternate burst samples have a value of 64.
DIGITAL ACTIVE LINE 948 SAMPLES (0–947)
TOTAL LINE 1135 SAMPLES (0–1134)
Figure 6.24. Digital Composite (B, D, G, H, I) PAL Analog and Digital Timing Relationship.
146
Chapter 6: Digital Video Interfaces
Ancillary Data
NTSC
PAL
Ancillary data packets are used to transmit information (such as digital audio, closed captioning, and teletext data) during the blanking intervals. ITU-R BT.1364 and SMPTE 291M describe the ancillary data formats. The ancillary data formats are the same as for digital component video, discussed earlier in this chapter. However, instead of a 3-word preamble, a one-word ancillary data flag is used, with a 10-bit value of 3FCH. There may be multiple ancillary data flags following the TRS-ID, with each flag identifying the beginning of another ancillary packet. Ancillary data may be present within the following word number boundaries (see Figures 6.25 through 6.30).
795–849 795–815 340–360 795–260 340–715
972–1035 972–994 404–426 972–302 404–869
END OF ANALOG LINE
horizontal sync period equalizing pulse periods vertical sync periods
User data may not use the 10-bit values of 000H–003H and 3FCH–3FFH, or the 8-bit values of 00H and FFH, since they are used for timing information.
25-pin Parallel Interface The SMPTE 244M parallel interface is based on that used for 27 MHz 4:2:2 digital component video (Table 6.19), except for the timing differences. This interface is used to transfer SDTV resolution digital composite
END OF DIGITAL LINE
768
782 784 50% 785
787
790–794
795–849
TRS–ID
ANC DATA (OPTIONAL)
Figure 6.25. (M) NTSC TRS-ID and Ancillary Data Locations During Horizontal Sync Intervals.
Pro-Video Composite Interfaces
50%
787
147
50%
790–794
795–260
TRS–ID
ANC DATA (OPTIONAL)
340–715 ANC DATA (OPTIONAL)
Figure 6.26. (M) NTSC TRS-ID and Ancillary Data Locations During Vertical Sync Intervals.
50%
787
50%
790–794
795–815
TRS–ID
ANC DATA (OPTIONAL)
340–360 ANC DATA (OPTIONAL)
Figure 6.27. (M) NTSC TRS-ID and Ancillary Data Locations During Equalizing Pulse Intervals.
148
Chapter 6: Digital Video Interfaces
END OF ANALOG LINE
END OF DIGITAL LINE
948
954 957 50% 958
962
967–971
972–1035
TRS–ID
ANC DATA (OPTIONAL)
Figure 6.28. (B, D, G, H, I) PAL TRS-ID and Ancillary Data Locations During Horizontal Sync Intervals.
50%
962
50%
967–971
972–302
404–869
TRS–ID
ANC DATA (OPTIONAL)
ANC DATA (OPTIONAL)
Figure 6.29. (B, D, G, H, I) PAL TRS-ID and Ancillary Data Locations During Vertical Sync Intervals.
Pro-Video Composite Interfaces
50%
962
149
50%
967–971
972–994
404–426
TRS–ID
ANC DATA (OPTIONAL)
ANC DATA (OPTIONAL)
Figure 6.30. (B, D, G, H, I) PAL TRS-ID and Ancillary Data Locations During Equalizing Pulse Intervals.
data. 8-bit or 10-bit data and a 4× FSC clock are transferred. Signal levels are compatible with ECLcompatible balanced drivers and receivers. The generator must have a balanced output with a maximum source impedance of 110 Ω; the signal must be 0.8–2.0V peak-to-peak measured across a 110-Ω load. At the receiver, the transmission line must be terminated by 110 ±10 Ω. The clock signal is a 4× FSC square wave, with a clock pulse width of 35 ±5 ns for (M) NTSC or 28 ±5 ns for (B, D, G, H, I) PAL. The positive transition of the clock signal occurs midway between data transitions with a tolerance of ±5 ns (as shown in Figure 6.31). To permit reliable operation at interconnect lengths of 50–200 meters, the receiver must use frequency equalization, with typical characteristics shown in Figure 6.3. This example enables operation with a range of cable lengths down to zero.
Serial Interface The parallel format can be converted to a SMPTE 259M serial format (Figure 6.32), allowing data to be transmitted using a 75-Ω coaxial cable (or optical fiber). This interface converts the 14.32 or 17.73 MHz parallel stream into a 143 or 177 Mbps serial stream. The 10× PLL generates the 143 or 177 MHz clock from the 14.32 or 17.73 MHz clock signal. For cable interconnect, the generator has an unbalanced output with a source impedance of 75Ω; the signal must be 0.8V ±10% peak-topeak measured across a 75-Ω load. The receiver has an input impedance of 75Ω. The 10 bits of data are serialized (LSB first) and processed using a scrambled and polarity-free NRZI algorithm: G(x) = (x9 + x4 + 1)(x + 1)
This algorithm is the same as used for digital component video discussed earlier. In an 8-bit
150
Chapter 6: Digital Video Interfaces
CLOCK TW TC
TD
DATA
TW = 35 ± 5 NS (M) NTSC; 28 ± 5 NS (B, D, G, H, I) PAL TC = 69.84 NS (M) NTSC; 56.39 NS (B, D, G, H, I) PAL TD = 35 ± 5 NS (M) NTSC; 28 ± 5 NS (B, D, G, H, I) PAL
Figure 6.31. Digital Composite Video Parallel Interface Waveforms.
TRS ID INSERTION
10–BIT DIGITAL COMPOSITE VIDEO
10
75-OHM COAX SHIFT REGISTER
10 SHIFT REGISTER
DESCRAMBLER
SCRAMBLER
4X FSC CLOCK
10–BIT DIGITAL COMPOSITE VIDEO
TRS–ID DETECT 10X PLL
40X FSC CLOCK 40X FSC PLL
DIVIDE BY 10
Figure 6.32. Serial Interface Block Diagram.
4X FSC CLOCK
Pro-Video Composite Interfaces
environment, 8-bit data is appended with two least significant “0” bits before serialization. The input signal to the scrambler (Figure 6.7) uses positive logic (the highest voltage represents a logical one; lowest voltage represents a logical zero). The formatted serial data is output at the 40× FSC rate. At the receiver, phase-lock synchronization is done by detecting the TRS-ID sequences. The PLL is continuously adjusted slightly each scan line to ensure that these patterns are detected and to avoid bit slippage. The recovered 10× clock is divided by ten to generate the 4× FSC sample clock. The serial data is lowand high-frequency equalized, inverse scrambling performed (Figure 6.8), and deserialized. TRS-ID When using the serial interface, a special five-word sequence, known as the TRS-ID, must be inserted into the digital video stream during the horizontal sync time. The TRS-ID is present only following sync leading edges which identify a horizontal transition, and
151
occupies horizontal counts 790–794, inclusive (NTSC) or 967–971, inclusive (PAL). Table 6.27 shows the TRS-ID format; Figures 6.25 through 6.30 show the TRS-ID locations for digital composite (M) NTSC and (B, D, G, H, I) PAL video signals. The line number ID word at horizontal count 794 (NTSC) or 971 (PAL) is defined as shown in Table 6.28. PAL requires the reset of the TRS-ID position relative to horizontal sync once per field on only one of lines 625–4 and 313–317 due to the 25-Hz offset. All lines have 1135 samples except the two lines used for reset, which have 1137 samples. The two additional samples are numbered 1135 and 1136, and occur just prior to the first active picture sample (sample 0). Due to the 25-Hz offset, the samples occur slightly earlier each line. Initial determination of the TRS-ID position should be done on line 1, Field 1, or a nearby line. The TRS-ID location always starts at sample 967, but the distance from the leading edge of sync varies due to the 25-Hz offset.
D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
TRS word 0
1
1
1
1
1
1
1
1
1
1
TRS word 1
0
0
0
0
0
0
0
0
0
0
TRS word 2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
D8
EP
TRS word 3 line number ID
line number ID
Notes: EP = even parity for D0–D7.
Table 6.27. TRS-ID Format.
152
Chapter 6: Digital Video Interfaces
D2
D1
D0
(M) NTSC
(B, D, G, H, I) PAL
0 0 0 0
0 0 1 1
0 1 0 1
line 1–263 field 1 line 264–525 field 2 line 1–263 field 3 line 264–525 field 4
line 1–313 field 1 line 314–625 field 2 line 1–313 field 3 line 314–625 field 4
1 1 1 1
0 0 1 1
0 1 0 1
not used not used not used not used
line 1–313 field 5 line 314–625 field 6 line 1–313 field 7 line 314–625 field 8
D7–D3
1 ≤ x ≤ 30 x = 31 x=0
(M) NTSC
(B, D, G, H, I) PAL
line number 1–30 [264–293] line number ≥ 31 [294] not used
line number 1–30 [314–343] line number ≥ 31 [344] not used
Table 6.28. Line Number ID Word at Horizontal Count 794 (NTSC) or 971 (PAL).
Pro-Video Transport Interfaces
such as SMPTE 326M. It may consist of MPEG-2 program or transport streams, DV streams, etc., and uses either 8-bit words plus even parity and D8, or 9-bit words plus D8.
Serial Data Transport Interface (SDTI) SMPTE 305M and ITU-R BT.1381 define a Serial Data Transport Interface (SDTI) that enables transferring data between equipment. The physical layer uses the 270 or 360 Mbps BT.656, BT.1302, and SMPTE 259M digital component video serial interface. Figure 6.33 illustrates the signal format. A 53-word header is inserted immediately after the EAV sequence, specifying the source, destination, and data format. Table 6.29 illustrates the header contents. The payload data is defined within BT.1381 and by other application-specific standards
Line Number The line number specifies a value of 1–525 (480i systems) or 1–625 (576i systems). L0 is the least significant bit. Line Number CRC The line number CRC applies to the data ID through the line number, for the entire 10 bits. C0 is the least significant bit. It is an 18-bit value, with an initial value set to all ones: CRC = x18 + x5 + x4 + x1
Pro-Video Transport Interfaces
E A V
HEADER
S A V
153
USER DATA (PAYLOAD)
Figure 6.33. SDTI Signal Format.
Code and AAI The 4-bit code value (CD3–CD0) specifies the length of the payload (the user data contained between the SAV and EAV sequences): 0000 0001 0010 1000
4:2:2 YCbCr video data 1440 word payload (uses 270 Mbps interface) 1920 word payload (uses 360 Mbps interface) 143 Mbps digital composite video
The 4-bit authorized address identifier (AAI) value, AAI3–AAI0, specifies the format of the destination and source addresses: 0000 0001
unspecified format IPv6 address
Destination and Source Addresses These specify the address of the source and destination devices. A universal address is indicated when all address bits are zero and AAI3–AAI0 = 0000.
Block Type The block type value specifies the segmentation of the payload. BL7–BL6 indicate the payload block structure: 00 01 10 11
fixed block size without ECC fixed block size with ECC unassigned variable block size
BL5–BL0 indicate the segmentation for fixed block sizes. Variable block sizes are indicated by BL7–BL0 having a value of 11000001. The ECC format is application-dependent. Payload CRC Flag The CRCF bit indicates whether or not the payload CRC is present at the end of the payload: 0 1
no CRC CRC present
154
Chapter 6: Digital Video Interfaces
10-bit Data D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
0
0
0
0
0
0
SDID
D8
EP
0
0
0
0
0
0
0
1
data count (DC)
D8
EP
0
0
1
0
1
1
1
0
D8
EP
L7
L6
L5
L4
L3
L2
L1
L0
D8
EP
0
0
0
0
0
0
L9
L8
D8
C8
C7
C6
C5
C4
C3
C2
C1
C0
D8
C17
C16
C15
C14
C13
C12
C11
C10
C9
D8
EP
AAI3
AAI2
AAI1
AAI0
CD3
CD2
CD1
CD0
D8
EP
DA7
DA6
DA5
DA4
DA3
DA2
DA1
DA0
D8
EP
DA15
DA14
DA13
DA12
DA11
DA10
DA9
DA8
ancillar y data flag (ADF)
line number
line number CRC code and AAI
destination address
source address
: D8
EP
DA127
DA126
DA125
DA124
DA123
DA122
DA121
DA120
D8
EP
SA7
SA6
SA5
SA4
SA3
SA2
SA1
SA0
D8
EP
SA15
SA14
SA13
SA12
SA11
SA10
SA9
SA8
SA124
SA123
SA122
SA121
SA120
: D8
EP
SA127
SA126
SA125
Notes: EP = even parity for D0–D7.
Table 6.29a. SDTI Header Structure.
Pro-Video Transport Interfaces
155
10-bit Data D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
block type
D8
EP
BL7
BL6
BL5
BL4
BL3
BL2
BL1
BL0
payload CRC flag
D8
EP
0
0
0
0
0
0
0
CRCF
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
D8
C8
C7
C6
C5
C4
C3
C2
C1
C0
D8
C17
C16
C15
C14
C13
C12
C11
C10
C9
header CRC
check sum
Sum of D0–D8 of data ID through last header CRC word. Preset to all zeros; carry is ignored.
D8
Notes: EP = even parity for D0–D7.
Table 6.29b. SDTI Header Structure (continued).
Header CRC The header CRC applies to the code and AAI word through the last reserved data word, for the entire 10 bits. C0 is the least significant bit. It is an 18-bit value, with an initial value set to all ones: CRC = x18 + x5 + x4 + x1
156
Chapter 6: Digital Video Interfaces
The payload data is defined by other application-specific standards. It may consist of MPEG-2 program or transport streams, DV streams, etc., and uses either 8-bit words plus even parity and D8, or 9-bit words plus D8.
High Data-Rate Serial Data Transport Interface (HD-SDTI) SMPTE 348M and ITU-R BT.1577 define a High Data-Rate Serial Data Transport Interface (HD-SDTI) that enables transferring data between equipment. The physical layer uses the 1.485 (or 1.485/1.001) Gbps SMPTE 292M digital component video serial interface. Figure 6.34 illustrates the signal format. Two data channels are multiplexed onto the single HD-SDTI stream such that one 74.25 (or 74.25/1.001) MHz data stream occupies the Y data space and the other 74.25 (or 74.25/1.001) MHz data stream occupies the CbCr data space. A 49-word header is inserted immediately after the line number CRC data, specifying the source, destination, and data format. Table 6.30 illustrates the header contents.
E A V
L N
C R C
HEADER
Code and AAI The 4-bit code value (CD3–CD0) specifies the length of the payload (the user data contained between the SAV and EAV sequences): 0000 0001 0010 0011 1000 1001 1010 1011 1100 1101 1110 1111
S A V
4:2:2 YCbCr video data 1440 word payload 1920 word payload 1280 word payload 143 Mbps digital composite video 2304 word payload (extended mode) 2400 word payload (extended mode) 1440 word payload (extended mode) 1728 word payload (extended mode) 2880 word payload (extended mode) 3456 word payload (extended mode) 3600 word payload (extended mode)
USER DATA (PAYLOAD)
C CHANNEL
E A V
L N
C R C
HEADER
S A V
USER DATA (PAYLOAD)
Y CHANNEL
Figure 6.34. HD-SDTI Signal Format. LN = line number (two 10-bit words), CRC = Line Number CRC (two 10-bit words).
Pro-Video Transport Interfaces
157
10-bit Data D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D8
EP
0
1
0
0
0
0
0
0
SDID
D8
EP
0
0
0
0
0
0
1
0
data count (DC)
D8
EP
0
0
1
0
1
0
1
0
code and AAI
D8
EP
AAI3
AAI2
AAI1
AAI0
CD3
CD2
CD1
CD0
D8
EP
DA7
DA6
DA5
DA4
DA3
DA2
DA1
DA0
D8
EP
DA15
DA14
DA13
DA12
DA11
DA10
DA9
DA8
ancillar y data flag (ADF)
destination address
source address
: D8
EP
DA127
DA126
DA125
DA124
DA123
DA122
DA121
DA120
D8
EP
SA7
SA6
SA5
SA4
SA3
SA2
SA1
SA0
D8
EP
SA15
SA14
SA13
SA12
SA11
SA10
SA9
SA8
: D8
EP
SA127
SA126
SA125
SA124
SA123
SA122
SA121
SA120
block type
D8
EP
BL7
BL6
BL5
BL4
BL3
BL2
BL1
BL0
payload CRC flag
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
Notes: EP = even parity for D0–D7.
Table 6.30a. HD-SDTI Header Structure.
158
Chapter 6: Digital Video Interfaces
10-bit Data D9 (MSB)
D8
D7
D6
D5
D4
D3
D2
D1
D0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
reser ved
D8
EP
0
0
0
0
0
0
0
0
D8
C8
C7
C6
C5
C4
C3
C2
C1
C0
D8
C17
C16
C15
C14
C13
C12
C11
C10
C9
header CRC
check sum
Sum of D0–D8 of data ID through last header CRC word. Preset to all zeros; carry is ignored.
D8
Notes: EP = even parity for D0–D7.
Table 6.30b. HD-SDTI Header Structure (continued).
The extended mode advances the timing of the SAV sequence, shortening the blanking interval, so that the payload data rate remains a constant 129.6 (or 129.6/1.001) MBps. The 4-bit authorized address identifier (AAI) format is the same as for SDTI. Destination and Source Addresses The source and destination address formats are the same as for SDTI.
Block Type The block type format is the same as for SDTI. Header CRC The header CRC applies to the DID through the last reserved data word, for the entire 10 bits. C0 is the least significant bit. It is an 18-bit value, with an initial value set to all ones: CRC = x18 + x5 + x4 + x1
IC Component Interfaces
IC Component Interfaces Many solutions for transferring digital video between chips are derived from the provideo interconnect standards. Chips for the pro-video market typically support 10 or 12 bits of data per video component, while chips for the consumer market typically use 8 bits of data per video component. “BT.601” and “BT.656” are the most popular interfaces for chips.
159
RGB Values: 8-bit Data Consumer solutions typically use 8-bit R´G´B´ data, with a range of 10H–EBH (note that PCs typically use a range of 00H–FFH). Values less than 10H or greater than EBH may be present due to processing. During blanking, R´G´B´ data should have a value of 10H, unless other information is present.
RGB Values: 10-bit Data YCbCr Values: 8-bit Data Y has a nominal range of 10H–EBH. Values less than 10H or greater than EBH may be present due to processing. Cb and Cr have a nominal range of 10H–F0H. Values less than 10H or greater than F0H may be present due to processing. YCbCr data may not use the values of 00H and FFH since those values may be used for timing information. During blanking, Y data should have a value of 10H and CbCr data should have a value of 80H, unless other information is present.
YCbCr Values: 10-bit Data For higher accuracy, pro-video solutions typically use 10-bit YCbCr data. Y has a nominal range of 040H–3ACH. Values less than 040H or greater than 3ACH may be present due to processing. Cb and Cr have a nominal range of 040H–3C0H. Values less than 040H or greater than 3C0H may be present due to processing. The values 000H–003H and 3FCH–3FFH may not be used to avoid timing contention with 8bit systems. During blanking, Y data should have a value of 040H and CbCr data should have a value of 200H, unless other information is present.
For higher accuracy, pro-video solutions typically use 10-bit R´G´B´ data, with a nominal range of 040H–3ACH. Values less than 040H or greater than 3ACH may be present due to processing. The values 000H–003H and 3FCH– 3FFH may not be used to avoid timing contention with 8-bit systems. During blanking, R´G´B´ data should have a value of 040H, unless other data is present.
“BT.601” Video Interface The “BT.601” video interface has been used for years, with the control signal names and timing reflecting the video standard. Supported active resolutions and sample clock rates are dependent on the video standard and aspect ratio. Devices usually support multiple data formats to simplify using them in a wide variety of applications. Video Data Formats The 24-bit 4:4:4 YCbCr data format is shown in Figure 6.35. Y, Cb, and Cr are each 8 bits, and all are sampled at the same rate, resulting in 24 bits of data per sample clock. Pro-video solutions typically use a 30-bit interface, with the Y, Cb, and Cr streams each being
160
Chapter 6: Digital Video Interfaces
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
Y 0
Y 1
Y 2
Y 3
Y 4
BLANKING
Y 5
Y 6
Y 7
Y [N - 1]
Y [N]
1 0
ACTIVE VIDEO
ONE SCAN LINE
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
C B 0
C B 1
C B 2
C B 3
C B 4
C B 5
C B 6
C B 7
C B [N - 1]
C B [N]
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
C R 0
C R 1
C R 2
C R 3
C R 4
C R 5
C R 6
C R 7
C R [N - 1]
C R [N]
8 0
Y [N - 1]
Y [N]
1 0
24-BIT 4:4:4 VIDEO
Figure 6.35. 24-Bit 4:4:4 YCbCr Data Format.
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
1 0
Y 0
Y 1
Y 2
Y 3
Y 4
BLANKING
Y 5
Y 6
Y 7
ACTIVE VIDEO 16-BIT 4:2:2 VIDEO
ONE SCAN LINE
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
C R 6
C C B R [N - 1] [N - 1]
8 0
Figure 6.36. 16-Bit 4:2:2 YCbCr Data Format.
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
BLANKING
C B 0
Y 0
C R 0
Y 1
C B 2
Y 2
C R 2
Y 3
ACTIVE VIDEO
ONE SCAN LINE
Figure 6.37. 8-Bit 4:2:2 YCbCr Data Format.
C R [N - 1]
Y [N]
8 0
8-BIT 4:2:2 VIDEO
IC Component Interfaces
10 bits. Y0, Cb0, and Cr0 are the least significant bits. The 16-bit 4:2:2 YCbCr data format is shown in Figure 6.36. Cb and Cr are sampled at one-half the Y sample rate, then multiplexed together. The CbCr stream of active data words always begins with a Cb sample. Provideo solutions typically use a 20-bit interface, with the Y and CbCr streams each being 10 bits. The 8-bit 4:2:2 YCbCr data format is shown in Figure 6.37. The Y and CbCr streams from the 16-bit 4:2:2 YCbCr format are simply multiplexed at 2× the sample clock rate. The YCbCr stream of active data words always begins with a Cb sample. Pro-video solutions typically use a 10-bit interface. Tables 6.31 and 6.32 illustrate the 15-bit RGB, 16-bit RGB, and 24-bit RGB formats. For the 15-bit RGB format, the unused bit is sometimes used for keying (alpha) information. R0, G0, and B0 are the least significant bits. Control Signals In addition to the video data, there are four control signals: HSYNC# VSYNC# BLANK# CLK
horizontal sync vertical sync blanking 1× or 2× sample clock
For the 8-bit and 10-bit 4:2:2 YCbCr data formats, CLK is a 2× sample clock. For the other data formats, CLK is a 1× sample clock. For sources, the control signals and video data are output following the rising edge of CLK. For receivers, the control signals and video data are sampled on the rising edge of CLK. While BLANK# is negated, active R´G´B´ or YCbCr video data is present. HSYNC# is asserted during the horizontal sync time each scan line, with the leading edge indicating the start of a new line. The amount
161
of time that HSYNC# is asserted is usually the same as that specified by the video standard. VSYNC# is asserted during the vertical sync time each field or frame, with the leading edge indicating the start of a new field or frame. The number of scan lines that VSYNC# is asserted is usually same as that specified by the video standard. For interlaced video, if the leading edges of VSYNC# and HSYNC# are coincident, the field is Field 1. If the leading edge of VSYNC# occurs mid-line, the field is Field 2. For noninterlaced video, the leading edge of VSYNC# indicates the start of a new frame. Figure 6.38 illustrates the typical HSYNC# and VSYNC# relationships. The 8-/10-bit interface is typically limited to SDTV resolutions. To support HDTV resolutions, some designs transfer data and control information using both edges of the clock. Receiver Considerations Assumptions should not be made about the number of samples per line or horizontal blanking interval. Otherwise, the implementation may not work with all sources. To ensure compatibility between various sources, horizontal counters should be reset by the leading edge of HSYNC#, not by the trailing edge of BLANK#. To handle real-world sources, a receiver should use a “window” for detecting whether Field 1 or Field 2 is present. For example, if the leading edge of VSYNC# occurs within ±64 1× clock cycles of the leading edge of HSYNC#, the field is Field 1. Otherwise, the field is Field 2. Some video sources indicate sync timing by having Y data be an 8-bit value less than 10H. However, most video ICs do not do this. In addition, to allow real-world video and test signals to be passed through with minimum disruption, many ICs now allow the Y data to
162
Chapter 6: Digital Video Interfaces
R7
24-bit 4:4:4 YCbCr Cr7
R6
Cr6
R5
Cr5
R4
Cr4
R3
Cr3
R2
Cr2
R1
Cr1
24-bit RGB
16-bit RGB (5,6,5)
15-bit RGB (5,5,5)
16-bit 4:2:2 YCbCr
8-bit 4:2:2 YCbCr
R0 G7
R4
–
Cr0 Y7
Y7
Cb7, Y7, Cr7
G6
R3
R4
Y6
Y6
Cb6, Y6, Cr6
G5
R2
R3
Y5
Y5
Cb5, Y5, Cr5
G4
R1
R2
Y4
Y4
Cb4, Y4, Cr4
G3
R0
R1
Y3
Y3
Cb3, Y3, Cr3
G2
G5
R0
Y2
Y2
Cb2, Y2, Cr2
G1
G4
G4
Y1
Y1
Cb1, Y1, Cr1
G0 B7
G3 G2
G3 G2
Y0 Cb7
Y0 Cb7, Cr7
Cb0, Y0, Cr0
B6
G1
G1
Cb6
Cb6, Cr6
B5
G0
G0
Cb5
Cb5, Cr5
B4
B4
B4
Cb4
Cb4, Cr4
B3
B3
B3
Cb3
Cb3, Cr3
B2
B2
B2
Cb2
Cb2, Cr2
B1
B1
B1
Cb1
Cb1, Cr1
B0
B0
B0
Cb0
Cb0, Cr0
Table 6.31. Transferring YCbCr and RGB Data over a 16-bit or 24-bit Interface.
IC Component Interfaces
16-bit RGB (5,6,5) R4
15-bit RGB (5,5,5) –
R3
R4
Y6
R2
R3
Y5
R1
R2
Y4
R0
R1
Y3
G5
R0
Y2
G4
G4
Y1
R7
G3 G2
G3 G2
Cr7
Y0 Cb7, Cr7
R6
G1
G1
Cr6
Cb6, Cr6
R5
G0
G0
Cr5
Cb5, Cr5
R4
B4
B4
Cr4
Cb4, Cr4
R3
B3
B3
Cr3
Cb3, Cr3
R2
B2
B2
Cr2
Cb2, Cr2
R1
B1
B1
Cr1
Cb1, Cr1
R0 G7
B0 R4
B0 –
Cr0 Y7
Cb0, Cr0 Y7
Cb7, Y7, Cr7
G6
R3
R4
Y6
Y6
Cb6, Y6, Cr6
G5
R2
R3
Y5
Y5
Cb5, Y5, Cr5
G4
R1
R2
Y4
Y4
Cb4, Y4, Cr4
G3
R0
R1
Y3
Y3
Cb3, Y3, Cr3
G2
G5
R0
Y2
Y2
Cb2, Y2, Cr2
G1
G4
G4
Y1
Y1
Cb1, Y1, Cr1
G0 B7
G3 G2
G3 G2
Y0 Cb7
Y0 Cb7, Cr7
Cb0, Y0, Cr0
B6
G1
G1
Cb6
Cb6, Cr6
B5
G0
G0
Cb5
Cb5, Cr5
B4
B4
B4
Cb4
Cb4, Cr4
B3
B3
B3
Cb3
Cb3, Cr3
B2
B2
B2
Cb2
Cb2, Cr2
B1
B1
B1
Cb1
Cb1, Cr1
B0
B0
B0
Cb0
Cb0, Cr0
24-bit RGB
24-bit 4:4:4 YCbCr
16-bit 4:2:2 YCbCr Y7
8-bit 4:2:2 YCbCr
Table 6.32. Transferring YCbCr and RGB Data over a 32-bit Interface.
163
164
Chapter 6: Digital Video Interfaces
START OF FIELD 1 OR FRAME
HSYNC#
VSYNC#
START OF FIELD 2
HSYNC#
VSYNC#
Figure 6.38. Typical HSYNC# and VSYNC# Relationships (Not to Scale).
have a value less than 10H during active video. Thus, receiver designs assuming sync timing is present on the Y channel may no longer work.
Video Module Interface (VMI) VMI (Video Module Interface) was developed in cooperation with several multimedia IC manufacturers. The goal was to standardize the video interfaces between devices such as MPEG decoders, NTSC/PAL decoders, and graphics chips.
Video Data Formats The VMI specification specifies an 8-bit 4:2:2 YCbCr data format as shown in Figure 6.39. Many devices also support the other YCbCr and R´G´B´ formats discussed in the “BT.601 Video Interface” section. Control Signals In addition to the video data, there are four control signals: HREF VREF VACTIVE PIXCLK
horizontal blanking vertical sync active video 2× sample clock
IC Component Interfaces
For the 8-bit and 10-bit 4:2:2 YCbCr data formats, PIXCLK is a 2× sample clock. For the other data formats, PIXCLK is a 1× sample clock. For sources, the control signals and video data are output following the rising edge of PIXCLK. For receivers, the control signals and video data are sampled on the rising edge of PIXCLK. While VACTIVE is asserted, active R´G´B´ or YCbCr video data is present. Although transitions in VACTIVE are allowed, it is intended to allow a hardware mechanism for cropping video data. For systems that do not support a VACTIVE signal, HREF can generally be connected to VACTIVE with minimal loss of function. To support video sources that do not generate a line-locked clock, a DVALID# (data valid) signal may also be used. While DVALID# is asserted, valid data is present. HREF is asserted during the active video time each scan line, including during the vertical blanking interval.
165
VREF is asserted for 6 scan line times, starting one-half scan line after the start of vertical sync. For interlaced video, the trailing edge of VREF is used to sample HREF. If HREF is asserted, the field is Field 1. If HREF is negated, the field is Field 2. For noninterlaced video, the leading edge of VREF indicates the start of a new frame. Figure 6.40 illustrates the typical HREF and VREF relationships. Receiver Considerations Assumptions should not be made about the number of samples per line or horizontal blanking interval. Otherwise, the implementation may not work with all sources. Video data has input setup and hold times, relative to the rising edge of PIXCLK, of 5 and 0 ns, respectively. VACTIVE has input setup and hold times, relative to the rising edge of PIXCLK, of 5 and 0 ns, respectively. HREF and VREF both have input setup and hold times, relative to the rising edge of PIXCLK, of 5 and 5 ns, respectively. HREF
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
8 0
1 0
BLANKING
C B 0
Y 0
C R 0
Y 1
C B 2
Y 2
C R 2
Y 3
C R [N - 1]
ACTIVE VIDEO
ONE SCAN LINE
Figure 6.39. VMI 8-bit 4:2:2 YCbCr Data for One Scan Line.
Y [N]
8 0
8-BIT 4:2:2 VIDEO
166
Chapter 6: Digital Video Interfaces
START OF FIELD 1 OR FRAME
HREF
VREF
START OF FIELD 2
HREF
VREF
Figure 6.40. VMI Typical HREF and VREF Relationships (Not to Scale).
“BT.656” Interface The BT.656 interface for ICs is based on the pro-video BT.656-type parallel interfaces, discussed earlier in this chapter (Figures 6.1 and 6.9). Using EAV and SAV sequences to indicate video timing reduces the number of pins required. The timing of the H, V, and F signals for common video formats is illustrated in Chapter 4. Standard IC signal levels and timing are used, and any resolution can be supported. Video Data Formats 8-bit or 10-bit 4:2:2 YCbCr data is used, as shown in Figures 6.1 and 6.9. Although
sources should generate the four protection bits in the EAV and SAV sequences, receivers may choose to ignore them due to the reliability of point-to-point transfers between chips. Control Signals CLK is a 2× sample clock. For sources, the video data is output following the rising edge of CLK. For receivers, the video data is sampled on the rising edge of CLK. This interface is typically limited to SDTV resolutions. To support HDTV resolutions, some designs transfer data using both edges of the clock.
IC Component Interfaces
Zoomed Video Port (ZV Port) Used on laptops, the ZV Port is a point-topoint uni-directional bus between the PC Card host adaptor and the graphics controller. It enables video data to be transferred real-time directly from the PC Card into the graphics frame buffer. The PC Card host adaptor has a special multimedia mode configuration. If a non-ZV PC Card is plugged into the slot, the host adaptor is not switched into the multimedia mode, and the PC Card behaves as expected. Once a ZV card has been plugged in and the host adaptor has been switched to the multimedia mode, the pin assignments change. As shown in Table 6.33, the PC Card signals A6–A25, SPKR#, INPACK#, and IOIS16# are replaced by ZV Port video signals (Y0–Y7, CbCr0– CbCr7, HREF, VREF, and PCLK) and 4-chan-
167
nel audio signals (MCLK, SCLK, LRCK, and SDATA). Video Data Formats 16-bit 4:2:2 YCbCr data is used, as shown in Figure 6.36. Control Signals In addition to the video data, there are four control signals: HREF VREF PCLK
horizontal reference vertical sync 1× sample clock
HREF, VREF, and PCLK have the same timing as the VMI interface discussed earlier in this chapter.
PC Card Signal
ZV Port Signal
PC Card Signal
ZV Port Signal
PC Card Signal
ZV Port Signal
A25
CbCr7
A17
Y1
A9
Y0
A24
CbCr5
A16
CbCr2
A8
Y2
A23
CbCr3
A15
CbCr4
A7
SCLK
A22
CbCr1
A14
Y6
A6
MCLK
A21
CbCr0
A13
Y4
SPKR#
SDATA
A20
Y7
A12
CbCr6
IOIS16#
PCLK
A19
Y5
A11
VREF
INPACK#
LRCK
A18
Y3
A10
HREF
Table 6.33. PC Card vs. ZV Port Signal Assignments.
168
Chapter 6: Digital Video Interfaces
Video Interface Port (VIP) The VESA VIP specification is an enhancement to the “BT.656” interface for ICs, previously discussed. The primary application is to interface up to four devices to a graphics controller chip, although the concept can easily be applied to other applications. There are three sections to the interface: Host Interface: VIPCLK HAD0–HAD7 HCTL
host clock host address/data bus host control
Video Interface: PIXCLK VID0–VID7 VIDA, VIDB
video sample clock lower video data bus 10-bit data extension
XPIXCLK XVID0–XVID7 XVIDA, XVIDB
video sample clock upper video data bus 10-bit data extension
System Interface: VRST# VIRQ#
reset interrupt request
The host interface signals are provided by the graphics controller. Essentially, a 2-, 4-, or 8-bit version of the PCI interface is used. VIPCLK has a frequency range of 25–33 MHz. PIXCLK and XPIXCLK have a maximum frequency of 75 and 80 MHz, respectively. Video Interface As with the “BT.656” interface, special four-word sequences are inserted into the 8-bit or 10-bit 4:2:2 YCbCr video stream to indicate the start of active video (SAV) and end of active video (EAV). These sequences also indicate
when horizontal and vertical blanking are present and which field is being transmitted. VIP modifies the BT.656 EAV and SAV sequences as shown in Table 6.34. BT.656 uses four protection bits (P0–P3) in the status word since it was designed for long cable connections between equipment. With chip-to-chip interconnect, this protection isn’t required, so the bits are used for other purposes. The timing of the H, V, and F signals for common video formats are illustrated in Chapter 4. The status word for VIP is defined as: T = “0” for task B T = “1” for task A F = “0” for Field 1 F = “1” for Field 2 V = “1” during vertical blanking H = “0” at SAV H = “1” at EAV
The task bit, T, is programmable. If BT.656 compatibility is required, it should always be a “1.” Otherwise, it may be used to indicate which one of two data streams are present: stream A = “1” and stream B = “0.” Alternately, T may be a “0” when raw 2× oversampled VBI data is present, and a “1” otherwise. The noninterlaced bit, N, indicates whether the source is progressive (“1”) or interlaced (“0”). The repeat bit, R, is a “1” if the current field is a repeat field. This occurs only during 3:2 pull-down. The repeat bit (R), in conjunction with the noninterlaced bit (N), enables the graphics controller to handle Bob and Weave, as well as 3:2 pull-down (further discussed in Chapter 7), in hardware. The extra flag bit, E, is a “1” if another byte follows the EAV. Table 6.35 illustrates the extra flag byte. This bit is valid only during EAV sequences. If the E bit in the extra byte is “1,” another extra byte immediately follows. This allows chaining any number of extra bytes together as needed.
IC Component Interfaces
169
8-bit Data D7 (MSB)
D6
D5
D4
D3
D2
D1
D0
1
1
1
1
1
1
1
1
preamble
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
status word
T
F
V
H
N
R
0
E
D1
D0
Table 6.34. VIP EAV and SAV Sequence.
8-bit Data D7 (MSB) extra byte
D6
D5
D4
D3
user defined
D0
D2
E
Table 6.35. VIP EAV Extra Byte.
Unlike pro-video interfaces, code 00H may be used during active video data to indicate an invalid video sample. This is used to accommodate scaled video and square pixel timing. Video Data Formats In the 8-bit mode (Figure 6.41), the video interface is similar to BT.656, except for the differences mentioned. XVID0–XVID7 are not used. In the 16-bit mode (Figure 6.42), SAV sequences, EAV sequences, Y video data, ancillary packet headers, and even-numbered ancillary data values are transferred across the lower 8 bits (VID0–VID7). CbCr video data and odd-numbered ancillary data values are transferred across the upper 8 bits (XVID0–XVID7).
Note that “skip data” (value 00H) during active video must also appear in 16-bit format to preserve the 16-bit data alignment. 10-bit video data is supported by the VIDA, VIDB, XVIDA, and XVIDB signals. VIDA and XVIDA are the least significant bits. Ancillary Data Ancillary data packets are used to transmit information (such as digital audio, closed captioning, and teletext data) during the blanking intervals, as shown in Table 6.36. Unlike provideo interfaces, the 00H and FFH values may be used by the ancillary data. Note that the ancillary data formats were defined prior to many of the pro-video ancillary data formats, and therefore may not match.
170
Chapter 6: Digital Video Interfaces
H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE F F
0 0
0 0
BLANKING X Y Z
8 0
1 0
8 0
SAV CODE
1 0
4
8 0
1 0
F F
0 0
268
0 0
CO–SITED X Y Z
C B 0
Y 0
NEXT LINE
CO–SITED
C R 0
Y 1
C B 2
Y 2
4
C R 2
Y 3
C R 718
Y 719
VIP 4:2:2 VIDEO
F F
1440
1716
Figure 6.41. VIP 8-Bit Interface Data for One Scan Line. 480i; 720 active samples per line; 27 MHz clock.
H SIGNAL
START OF DIGITAL LINE
START OF DIGITAL ACTIVE LINE
EAV CODE F F
0 0
0 0
BLANKING X Y Z
1 0
1 0
1 0
4
1 0
NEXT LINE
SAV CODE 1 0
1 0
F F
0 0
272
0 0
X Y Z
Y 0
Y 1
Y 2
Y 3
Y 4
Y 5
4
Y 6
Y 7
Y 1918
Y 1919
F F
C R 6
C B 1918
C R 1918
F F
Y CHANNEL
1920
2200
BLANKING 8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
8 0
C B 0
C R 0
C B 2
C R 2
C B 4
C R 4
C B 6
Figure 6.42. VIP 16-Bit Interface Data for One Scan Line. 1080i; 1920 active samples per line; 74.176 or 74.25 MHz clock.
CBCR CHANNEL
IC Component Interfaces
171
8-bit Data D7 (MSB)
D6
D5
D4
D3
D2
D1
D0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
data ID (DID)
D6
EP
0
1
0
DID2
DID1
DID0
SDID
D6
EP
data count (DC)
D6
EP
DC2
DC1
DC0
D3
D2
D1
D0
ancillar y data flag (ADF)
user defined value DC5
DC4
DC3
internal data ID 0
user defined value
internal data ID 1
user defined value D7
data word 0
D6
D5
D4 :
: data word N
D7
D6
D5
D4
D3
D2
D1
D0
check sum
D6
EP
CS5
CS4
CS3
CS2
CS1
CS0
optional fill data
D6
EP
0
0
0
0
0
0
Notes: EP = even parity for D0–D5.
Table 6.36. VIP Ancillary Data Packet General Format.
DID2 of the DID field indicates whether Field 1 or Field 2 ancillary data is present: 0 1
Field 1 Field 2
DID1–DID0 of the DID field indicate the type of ancillary data present: 00 01 10 11
start of field sliced VBI data, lines 1–23 end of field VBI data, line 23 sliced VBI data, line 24 to end of field
The data count value (DC) specifies the number of D-words (4-byte blocks) of ancillary data present. Thus, the number of data words in the ancillary packet after the DID must be a multiple of four. 1–3 optional fill bytes may be added after the check sum data to meet this requirement. When DID1–DID0 are “00” or “10,” no ancillary data or check sum is present. The data count (DC) value is “00000,” and is the last field present in the packet.
172
Chapter 6: Digital Video Interfaces
Consumer Component Interfaces Many solutions for transferring digital video between equipment has been developed over the years. The DVI and HDMI standards are the most popular digital video interfaces for consumer equipment.
Digital Visual Interface (DVI) In 1998, the Digital Display Working Group (DDWG) was formed to address the need for a standardized digital video interface between a PC and VGA monitor, as illustrated in Figure 6.43. The DVI 1.0 specification was released in April 1999. Designed to transfer uncompressed realtime digital video, DVI supports PC graphics resolutions beyond 1600 × 1200 and HDTV resolutions, including 720p, 1080i, and 1080p. In 2003, the consumer electronics industry started adding DVI outputs to DVD players and cable/satellite settop boxes. DVI inputs also started appearing on digital televisions and LCD/plasma monitors.
Technology DVI is based on the Digital Flat Panel (DFP) Interface, enhancing it by supporting more formats and timings. It also includes support for the High-bandwidth Digital Content Protection (HDCP) specification to deter unauthorized copying of content. DVI also supports VESA’s Extended Display Identification Data (EDID) standard, Display Data Channel (DDC) standard (used to read the EDID), and Monitor Timing Specification (DMT). DDC and EDID enable automatic display detection and configuration. Extended Display Identification Data (EDID) was created to enable plug and play capabilities of displays. Data is stored in the display, describing the supported video formats This information is supplied to the source device, over DVI, at the request of the source device. The source device then chooses its output format, taking into account the format of the original video stream and the formats supported by the display. The source device is responsible for the format conversions necessary to supply video in an understandable form to the display.
R G PC
B
MONITOR
PC
DVI
MONITOR
H V
WITHOUT DVI
WITH DVI
Figure 6.43. Using DVI to Connect a VGA Monitor to a PC.
Consumer Component Interfaces
In addition, the EIA-861 standard specifies mandatory and optionally supported resolutions and timings, and how to include data such as aspect ratio and format information. TMDS Links DVI uses transition-minimized differential signaling (TMDS). Eight bits of video data are converted to a 10-bit transition-minimized, DCbalanced value, which is then serialized. The receiver deserializes the data, and converts it back to eight bits. Thus, to transfer digital R´G´B´ data requires three TMDS signals that comprise one TMDS link. “TFT data mapping” is supported as the minimum requirement: one pixel per clock, eight bits per channel, MSB justified. Either one or two TMDS links may be used, as shown in Figures 6.44 and 6.45, depending on the formats and timing required. A system supporting two TMDS links must be able to switch dynamically between formats requiring a single link and formats requiring a dual link. A single DVI connector can handle two TMDS links. TMDS TRANSMITTER
TMDS LINK
ENCODER AND SERIALIZER
CHANNEL 0
A single TMDS link supports resolutions and timings using a video sample rate of 25– 165 MHz. Resolutions and timings using a video sample rate of 165–330 MHz are implemented using two TMDS links, with each TMDS link operating at one-half the frequency. Thus, the two TMDS links share the same clock and the bandwidth is shared evenly between the two links. Video Data Formats Typically, 24-bit R´G´B´ data is transferred over a link. For applications requiring more than eight bits per color component, the second TMDS link may be used for the additional least significant bits. For PC applications, R´G´B´ data typically has a range of 00H–FFH. For consumer applications, R´G´B´ data typically has a range of 10H– EBH (values less than 10H or greater than EBH may be occasionally present due to processing).
TMDS RECEIVER
B0–B7
B0–B7 VSYNC HSYNC
RECEIVER AND DECODER
DE
CTL1
ENCODER AND SERIALIZER
CHANNEL 1
RECEIVER AND DECODER
CTL3
CLK
VSYNC
HSYNC
HSYNC
DE0
DE
G0–G7
CTL0 CTL1 DE1
CTL0 INTER CHANNEL ALIGNMENT
R0–R7
R0–R7 CTL2
B0–B7
VSYNC
G0–G7
G0–G7 CTL0
ENCODER AND SERIALIZER
CHANNEL 2
173
RECEIVER AND DECODER
CTL2 CTL3
CTL1
R0–R7 CTL2 CTL3
DE2
CHANNEL C
Figure 6.44. DVI Single TMDS Link.
CLK
174
Chapter 6: Digital Video Interfaces
TMDS TRANSMITTER
DUAL TMDS LINK
TMDS RECEIVER
B0–B7
B0–B7 VSYNC HSYNC
ENCODER AND SERIALIZER
CHANNEL 0
RECEIVER AND DECODER
DE
CTL1
ENCODER AND SERIALIZER
CHANNEL 1
RECEIVER AND DECODER
CTL3
ENCODER AND SERIALIZER
CHANNEL 2
RECEIVER AND DECODER
CTL7
CHANNEL 3
ENCODER AND SERIALIZER
CHANNEL 4
RECEIVER AND DECODER
CTL9
CTL0
CTL1
CTL1
DE1
R0–R7
CTL2
CTL2
CTL3
CTL3
DE2
CTL4 CTL5
CTL6
CHANNEL 5
RECEIVER AND DECODER
B0–B7 CTL4 CTL5
CTL7
G0–G7 CTL6 CTL7
DE4
R0–R7 ENCODER AND SERIALIZER
CLK
DE3
G0–G7 RECEIVER AND DECODER
R0–R7 CTL8
G0–G7
CTL0
B0–B7 ENCODER AND SERIALIZER
G0–G7 CTL6
DE
INTER CHANNEL ALIGNMENT
B0–B7 CTL5
HSYNC
DE0
CHANNEL C
CLK
CTL4
HSYNC
R0–R7
R0–R7 CTL2
VSYNC
G0–G7
G0–G7 CTL0
B0–B7
VSYNC
CTL8 CTL9 DE5
Figure 6.45. DVI Dual TMDS Link.
R0–R7 CTL8 CTL9
Consumer Component Interfaces
Control Signals In addition to the video data, DVI transmitter and receiver chips typically use up to 14 control signals for interfacing to other chips in the system: HSYNC VSYNC DE CTL0–CTL3 CTL4–CTL9 CLK
horizontal sync vertical sync data enable reserved (link 0) reserved (link 1) 1× sample clock
While DE is a “1,” active video is processed. While DE is a “0,” the HSYNC, VSYNC and CTL0–CTL9 signals are processed. HSYNC and VSYNC may be either polarity. One issue is that some HDTVs use the falling edge of the YPbPr tri-level sync, rather than the center (rising edge), for horizontal timing. When displaying content from DVI, this results in the image shifting by 2.3%. Providing the ability to adjust the DVI embedded sync timing relative to the YPbPr tri-level sync timing is a useful capability in this case. Many fixed-pixel displays, such as DLP, LCD and plasma, instead use the DE signal as a timing reference, avoiding the issue.
Digital-Only (DVI-D) Connector The digital-only connector, which supports dual link operation, contains 24 contacts arranged as three rows of eight contacts, as shown in Figure 6.46. Table 6.37 lists the pin assignments. Digital-Analog (DVI-I) Connector In addition to the 24 contacts used by the digital-only connector, the 29-contact digitalanalog connector adds five additional contacts to support analog video as shown in Figure 6.47. Table 6.38 lists the pin assignments. HSYNC VSYNC RED GREEN BLUE
horizontal sync vertical sync analog red video analog green video analog blue video
The operation of the analog signals is the same as for a standard VGA connector. DVI-A is available as a plug (male) connector only and mates to the analog-only pins of a DVI-I connector. DVI-A is only used in adapter cables, where there is the need to convert to or from a traditional analog VGA signal.
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
17
18
19
20
21
22
23
24
C1
C2
C3
C4 C5
Figure 6.46. DVI-D Connector.
175
Figure 6.47. DVI-I Connector.
176
Chapter 6: Digital Video Interfaces
Pin
Signal
Pin
Signal
Pin
SIgnal
1
D2–
9
D1–
17
D0–
2
D2
10
D1
18
D0
3
shield
11
shield
19
shield
4
D4–
12
D3–
20
D5–
5
D4
13
D3
21
D5
6
DDC SCL
14
+5V
22
shield
7
DDC SDA
15
ground
23
CLK
8
reserved
16
Hot Plug Detect
24
CLK–
Table 6.37. DVI-D Connector Signal Assignments.
Pin
Signal
Pin
Signal
Pin
SIgnal
1
D2–
9
D1–
17
D0–
2
D2
10
D1
18
D0
3
shield
11
shield
19
shield
4
D4–
12
D3–
20
D5–
5
D4
13
D3
21
D5
6
DDC SCL
14
+5V
22
shield
7
DDC SDA
15
ground
23
CLK
8
VSYNC
16
Hot Plug Detect
24
CLK–
C1
RED
C2
GREEN
C3
BLUE
C4
HSYNC
C5
ground
Table 6.38. DVI-I Connector Signal Assignments.
Consumer Component Interfaces
High Definition Multimedia Interface (HDMI) Although DVI handles transferring uncompressed real-time digital RGB video to a display, the consumer electronics industry preferred a smaller, more flexible solution, based on DVI technology. In April 2002, the HDMI working group was formed by Hitachi, Matsushita Electric (Panasonic), Philips, Silicon Image, Sony, Thomson and Toshiba. HDMI is capable of replacing up to eight audio cables (7.1 channels) and up to three video cables with a single cable, as illustrated in Figure 6.48. In 2004, the consumer electronics industry started adding HDMI outputs to DVD players and cable/satellite set-top boxes. HDMI inputs also started appearing on digital televisions and LCD/plasma monitors, with wide usage expected by 2005. Through the use of an adaptor cable, HDMI is backwards compatible with equipment using DVI and the EIA-861 DTV profile. However, the advanced features of HDMI, such as digital audio and Consumer Electronics Control (used to enable passing control commands between equipment), are not available.
DVD PLAYER OR SET-TOP BOX
Y PR AUDIO
WITHOUT HDMI
Technology HDMI, based on DVI, supports VESA’s Extended Display Identification Data (EDID) standard and Display Data Channel (DDC) standard (used to read the EDID). In addition, the EIA-861 standard specifies mandatory and optionally supported resolutions and timings, and how to include data such as aspect ratio and format information HDMI also supports the High-bandwidth Digital Content Protection (HDCP) specification to deter unauthorized copying of content. A common problem is sources not polling the TV often enough (twice per second) to see if its HDCP circuit is active. This results in “snow” if the TV’s HDMI input is deselected, then later selected again. The 19-pin “Type A” connector uses a single TMDS link and can therefore carry video signals with a 25–165 MHz sample rate. Video with sample rates below 25 MHz (i.e. 13.5 MHz 480i and 576i) are transmitted using a pixelrepetition scheme. This smaller connector is well-suited for consumer devices due to its small size. To support video signals sampled at greater than 165 MHz, the dual-link capability of the 29-pin “Type B” connector designed for the PC market must be used.
DVD
PB DISPLAY
177
PLAYER OR SET-TOP BOX
HDMI
DISPLAY
WITH HDMI
Figure 6.48. Using HDMI Eliminates Confusing Cable Connections For Consumers.
178
Chapter 6: Digital Video Interfaces
Video Data Formats HDMI supports RGB, 4:4:4 YCbCr and 4:2:2 YCbCr. Up to 24 bits per pixel can be transferred. For the 640 × 480 resolution, the R´G´B´ data has a range of 00H–FFH. For YCbCr data and all other RGB resolutions, data has a range of 10H–EBH (values less than 10H or greater than EBH may be occasionally present due to processing). Audio Data Formats Driven by the DVD-Audio standard, audio support consists of 1–8 uncompressed audio streams with a sample rate of up to 48, 96 or 192 kHz, depending on the video format. It can alternately carry a compressed multi-channel audio stream at sample rates up to 192 kHz.
TMDS TRANSMITTER
TMDS LINK
ENCODER AND SERIALIZER
CHANNEL 0
Digital Flat Panel (DFP) Interface The VESA DFP interface was developed for transferring uncompressed digital video from a computer to a digital flat panel display. It supports VESA’s Plug and Display (P&D) standard, Extended Display Identification Data (EDID) standard, Display Data Channel (DDC) standard, and Monitor Timing Specification (DMT). DDC and EDID enable automatic display detection and configuration. Only “TFT data mapping” is supported: one pixel per clock, eight bits per channel, MSB justified. Like DVI, DFP uses transition-minimized differential signaling (TMDS). Eight bits of video data are converted to a 10-bit transitionminimized, DC-balanced value, which is then serialized. The receiver deserializes the data, and converts it back to eight bits. Thus, to transfer digital R´G´B´ data requires three TMDS signals that comprise one TMDS link. Cable lengths may be up to 5 meters.
TMDS RECEIVER
B0–B7
B0–B7 VSYNC HSYNC
RECEIVER AND DECODER
DE
CTL1
ENCODER AND SERIALIZER
CHANNEL 1
RECEIVER AND DECODER
CTL3
CLK
HSYNC
DE0
DE
ENCODER AND SERIALIZER
CHANNEL 2
RECEIVER AND DECODER
G0–G7
CTL0 CTL1 DE1
R0–R7
R0–R7 CTL2
VSYNC
HSYNC
G0–G7
G0–G7 CTL0
B0–B7
VSYNC
CTL2 CTL3
CTL0 INTER CHANNEL ALIGNMENT
CTL1
R0–R7 CTL2 CTL3
DE2
CHANNEL C
Figure 6.49. DFP TMDS Link.
CLK
Consumer Component Interfaces
10
9
8
7
6
5
4
3
2
1
20
19
18
17
16
15
14
13
12
11
179
Figure 6.50. DFP Connector.
TMDS Links A single TMDS link, as shown in Figure 6.49, supports formats and timings requiring a clock rate of 22.5–160 MHz. Video Data Formats 24-bit R´G´B´ data is transferred over the link, as shown in Figure 6.49. Control Signals In addition to the video data, DFP transmitter and receiver chips typically use up to 8 control signals for interfacing to other chips in the system: HSYNC VSYNC DE CTL0–CTL3 CLK
horizontal sync vertical sync data enable reserved 1× sample clock
While DE is a “1,” active video is processed. While DE is a “0,” the HSYNC, VSYNC and CTL0–CTL3 signals are processed. HSYNC and VSYNC may be either polarity. Connector The 20-pin mini-D ribbon (MDR) connector contains 20 contacts arranged as two rows of ten contacts, as shown in Figure 6.50. Table 6.39 lists the pin assignments.
Pin
Signal
Pin
Signal
1
D1
11
D2
2
D1–
12
D2–
3
shield
13
shield
4
shield
14
shield
5
CLK
15
D0
6
CLK–
16
D0–
7
ground
17
no connect
8
+5V
18
Hot Plug Detect
9
no connect
19
DDC SDA
10
no connect
20
DDC SCL
Table 6.39. DFP Connector Signal Assignments.
180
Chapter 6: Digital Video Interfaces
Open LVDS Display Interface (OpenLDI) OpenLDI was developed for transferring uncompressed digital video from a computer to a digital flat panel display. It enhances the FPD-Link standard used to drive the displays of laptop computers, and adds support for VESA’s Plug and Display (P&D) standard, Extended Display Identification Data (EDID) standard, and Display Data Channel (DDC) standard. DDC and EDID enable automatic display detection and configuration. Unlike DVI and DFP, OpenLDI uses lowvoltage differential signaling (LVDS). Cable lengths may be up to 10 meters. LVDS Link The LVDS link, as shown in Figure 6.51, supports formats and timings requiring a clock rate of 32.5–160 MHz. Eight serial data lines (A0–A7) and two sample clock lines (CLK1 and CLK2) are used. The number of serial data lines actually used is dependent on the pixel format, with the serial data rate being 7× the sample clock rate. The CLK2 signal is used in the dual pixel modes for backwards compatibility with FPD-Link receivers. Video Data Formats 18-bit single pixel, 24-bit single pixel, 18-bit dual pixel, or 24-bit dual pixel R´G´B´ data is transferred over the link. Table 6.40 illustrates the mapping between the pixel data bit number and the OpenLDI bit number. The 18-bit single pixel R´G´B´ format uses three 6-bit R´G´B´ values: R0–R5, G0–G5, and B0–B5. OpenLDI serial data lines A0–A2 are used to transfer the data. The 24-bit single pixel R´G´B´ format uses three 8-bit R´G´B´ values: R0–R7, G0–G7, and
B0–B7. OpenLDI serial data lines A0–A3 are used to transfer the data. The 18-bit dual pixel R´G´B´ format represents two pixels as three upper/lower pairs of 6-bit R´G´B´ values: RU0–RU5, GU0–GU5, BU0–BU5, RL0–RL5, GL0–GL5, BL0–BL5. Each upper/lower pair represents two pixels. OpenLDI serial data lines A0–A2 and A4–A6 are used to transfer the data. The 24-bit dual pixel R´G´B´ format represents two pixels as three upper/lower pairs of 8-bit R´G´B´ values: RU0–RU7, GU0–GU7, BU0–BU7, RL0–RL7, GL0–GL7, BL0–BL7. Each upper/lower pair represents two pixels. OpenLDI serial data lines A0–A7 are used to transfer the data. Control Signals In addition to the video data, OpenLDI transmitter and receiver chips typically use up to seven control signals for interfacing to other chips in the system: HSYNC VSYNC DE CNTLE CNTLF CLK1 CLK2
horizontal sync vertical sync data enable reserved reserved 1× sample clock 1× sample clock
During unbalanced operation, the DE, HSYNC, VSYNC, CNTLE, and CNTLF levels are sent as unencoded bits within the A2 and A6 bitstreams. During balanced operation (used to minimize short- and long-term DC bias), a DC Balance bit is sent within each of the A0–A7 bitstreams to indicate whether the data is unmodified or inverted. Since there is no room left for the control signals to be sent directly, the DE level is sent by slightly modifying the
Consumer Component Interfaces
LVDS TRANSMITTER
LVDS LINK
LVDS RECEIVER
B0–B7
B0–B7 CHANNEL 0
VSYNC HSYNC
VSYNC HSYNC
CHANNEL 1
DE
DE
CHANNEL 2
G0–G7 ENCODER AND SERIALIZER
CNTLE
CHANNEL 3 CHANNEL 4
G0–G7 RECEIVER AND DECODER
CNTLE
CHANNEL 5
R0–R7
R0–R7
CHANNEL 6
CNTLF
CNTLF
CHANNEL 7
CLK1
CLK1
CLK2
CLK2
Figure 6.51. OpenLDI LVDS Link.
18 Bits per Pixel Bit Number
24 Bits per Pixel Bit Number
OpenLDI Bit Number
5
7
5
4
6
4
3
5
3
2
4
2
1
3
1
0
2
0
1
7
0
6
Table 6.40. OpenLDI Bit Number Mappings.
181
182
Chapter 6: Digital Video Interfaces
timing of the falling edge of the CLK1 and CLK2 signals. The HSYNC, VSYNC, CNTLE and CNTLF levels are sent during the blanking intervals using 7-bit code words on the A0, A1, A5, and A4 signals, respectively. Connector The 36-pin mini-D ribbon (MDR) connector is similar to the one shown in Figure 6.50, except that there are two rows of eighteen contacts. Table 6.41 lists the pin assignments.
Gigabit Video Interface (GVIF) The Sony GVIF was developed for transferring uncompressed digital video using a single differential signal, instead of the multiple signals that DVI, DFP, and OpenLDI use. Cable lengths may be up to 10 meters.
GVIF Link The GVIF link, as shown in Figure 6.52, supports formats and timings requiring a clock rate of 20–80 MHz. For applications requiring higher clock rates, more than one GVIF link may be used. The serial data rate is 24× the sample clock rate for 18-bit R´G´B´ data, or 30× the sample clock rate for 24-bit R´G´B´ data. Video Data Formats 18-bit or 24-bit R´G´B´ data, plus timing, is transferred over the link. The 18-bit R´G´B´ format uses three 6-bit R´G´B´ values: R0–R5, G0– G5, and B0–B5. The 24-bit R´G´B´ format uses three 8-bit R´G´B´ values: R0–R7, G0–G7, and B0–B7.
Pin
Signal
Pin
Signal
Pin
Signal
1
A0–
13
+5V
25
reserved
2
A1–
14
A4–
26
reserved
3
A2–
15
A5–
27
ground
4
CLK1–
16
A6–
28
DDC SDA
5
A3–
17
A7–
29
ground
6
ground
18
CLK2–
30
USB–
7
reserved
19
A0
31
ground
8
reserved
20
A1
32
A4
9
reserved
21
A2
33
A5
10
DDC SCL
22
CLK1
34
A6
11
+5V
23
A3
35
A7
12
USB
24
reserved
36
CLK2
Table 6.41. OpenLDI Connector Signal Assignments.
Consumer Component Interfaces
18-bit R´G´B´ data is converted to 24-bit data by slicing the R´G´B data into six 3-bit values that are in turn transformed into six 4-bit codes. This ensures rich transitions for receiver PLL locking and good DC balance. 24-bit R´G´B´ data is converted to 30-bit data by slicing the R´G´B data into six 4-bit values that are in turn transformed into six 5-bit codes.
CTL0 CTL1 CLK
horizontal sync vertical sync data enable
GVIF TRANSMITTER
reserved reserved 1× sample clock
If any of the HSYNC, VSYNC, DE, CTL0, or CTL1 signals change, during the next CLK cycle a special 30-bit format is used. The first six bits are header data indicating the new levels of HSYNC, VSYNC, DE, CTL0, or CTL1. This is followed by 24 bits of R´G´B data (unencoded except for inverting the odd bits). Note that during the blanking periods, non-video data, such as digital audio, may be transferred. The CTL signals may be used to indicate when non-video data is present.
Control Signals In addition to the video data, there are six control signals: HSYNC VSYNC DE
GVIF LINK
GVIF RECEIVER
B0–B7
B0–B7
VSYNC
VSYNC
HSYNC
HSYNC
DE
DE
G0–G7
G0–G7 CTL0
CTL0 ENCODER AND SERIALIZER
183
SDATA
RECEIVER AND DECODER
R0–R7
R0–R7
CTL1
CTL1
CLK
CLK
Figure 6.52. GVIF Link.
184
Chapter 6: Digital Video Interfaces
Consumer Transport Interfaces Several transport interfaces, such as USB 2.0, Ethernet and IEEE 1394, are available for consumer products. Of course, each standard has its own advantages and disadvantages.
USB 2.0 Well known in the PC market for connecting peripherals to a PC, there is growing interest in using USB (Universal Serial Bus) 2.0 to transfer compressed audio/video data between products. USB 2.0 is capable of operating up to 480 Mbps and supports an isochronous mode to guarantee data delivery timing. Thus, it can easily transfer compressed real-time audio/ video data from a cable/satellite set-top box or DVD player to a digital television. DTCP (Digital Transmission Copy Protection) may be used to encrypt the audio and video content over USB. Due to USB’s lower cost and widespread usage, many companies are interested in using USB 2.0 instead of IEEE 1394 to transfer compressed audio/video data between products. However, some still prefer IEEE 1394 since the methods for transferring various types of data are much better defined. USB On-The-Go With portable devices increasing in popularity, there was a growing desire for them to communicate directly with each other without requiring a PC or other USB host. “On-The-Go” addresses this desire by allowing a USB device to communicate directly with other “On-The-Go” products. It also features a smaller USB connector and low power features to preserve battery life.
Ethernet With the widespread adoption of home networks, DSL and FTTH (Fiber-To-The-Home), Ethernet has become a common interface for transporting digital audio and video data. Initially used for file transfers, streaming of realtime compressed video over wired or wireless Ethernet networks is now becoming common. Ethernet commonly supports up to 100 Mbps, with use of 1 Gbps starting to increase. DTCP (Digital Transmission Copy Protection) may be used to encrypt the audio and video content over wired or wireless networks.
IEEE 1394 IEEE 1394 was originally developed by Apple Computer as Firewire. Designed to be a generic interface between devices, 1394 specifies the physical characteristics; separate application-specific specifications describe how to transfer data over the 1394 network. The OpenCable™, SCTE DVS-194, EIA-775, and ITU-T J.117 specifications for compatibility between digital televisions and settop boxes specifically include IEEE 1394 support. 1394 is a transaction-based packet technology, using a bi-directional serial interconnect that features hot plug-and-play. This enables devices to be connected and disconnected without affecting the operation of other devices connected to the network. Guaranteed delivery of time-sensitive data is supported, enabling digital audio and video to be transferred in real time. In addition, multiple independent streams of digital audio and video can be carried. Specifications The original 1394-1995 specification supports bit rates of 98.304, 196.608, and 393.216 Mbps.
Consumer Transport Interfaces
185
16 HOPS = 17 NODES MAX.
1
2
3
4
16
17
BRANCHING INCREASES NODE COUNT
1
2
3
4
16
18
19
17
20
21
Figure 6.53. IEEE 1394 Network Topology Examples.
The 1394A-2000 specification clarifies areas that were vague and led to system interoperability issues. It also reduces the overhead lost to bus control, arbitration, bus reset duration, and concatenation of packets. 1394A-2000 also introduces advanced powersavings features. The electrical signalling method is also common between 1394-1995 and 1394A-2000, using data-strobe (DS) encoding and analog-speed signaling. The 1394B-2002 specification adds support for bit rates of 786.432, 1572.864, and 3145.728 Mbps. It also includes - 8B/10B encoding technique used by Gigabit Ethernet - Continuous dual simplex operation - Longer distance (up to 100 meters over Cat5) - Changes the speed signalling to a more digital method
- Three types of ports: Legacy (1395A compatible), Beta and Bilingual (supports both Legacy and Beta). Connector keying ensures that incompatible connections cannot physically be made.
Endian Issues 1394 uses a big-endian architecture, defining the most significant bit as bit 0. However, many processors are based on the little endian architecture which defines the most significant bit as bit 31 (assuming a 32-bit word). Network Topology Like many networks, there is no designated bus master. The tree-like network structure has a root node, branching out to logical nodes in other devices (Figure 6.53). The root is responsible for certain control functions, and is chosen during initialization. Once chosen, it retains that function for as long as it remains powered-on and connected to the network.
186
Chapter 6: Digital Video Interfaces
A network can include up to 63 nodes, with each node (or device) specified by a 6-bit physical identification number. Multiple networks may be connected by bridges, up to a system maximum of 1,023 networks, with each network represented by a separate 10-bit bus ID. Combined, the 16-bit address allows up to 64,449 nodes in a system. Since device addresses are 64 bits, and 16 of these bits are used to specify nodes and networks, 48 bits remain for memory addresses, allowing up to 256TB of memory space per node. Node Types Nodes on a 1394 bus may vary in complexity and capability (listed simplest to most complex): Transaction nodes respond to asynchronous communication, implement the minimal set of control status registers (CSR), and implement a minimal configuration ROM. Isochronous nodes add a 24.576 MHz clock used to increment a cycle timer register that is updated by cycle start packets. Cycle master nodes add the ability to generate the 8 kHz cycle start event, generate cycle start packets, and implement a bus timer register.
Isochronous resource manager (IRM) nodes add the ability to detect bad self-ID packets, determine the node ID of the chosen IRM, and implement the channels available, bandwidth available, and bus manager ID registers. At least one node must be capable of acting as an IRM to support isochronous communication. Bus manager (BM) nodes are the most complex. This level adds responsibility for storing every self-ID packet in a topology map and analyzing that map to produce a speed map of the entire bus. These two maps are used to manage the bus. Finally, the BM must be able to activate the cycle master node, write configuration packets to allow optimization of the bus, and act as the power manager. Node Ports In the network topology, a one-port device is known as a “leaf” device since it is at the end of a network branch. They can be connected to the network, but cannot expand the network. Two-port devices can be used to form daisy-chained topologies. They can be connected to and continue the network, as shown in Figure 6.53. Devices with three or more ports are able to branch the network to the full 63-node capability.
DATA
STROBE
STROBE XOR DATA
Figure 6.54. IEEE 1394 Data and Strobe SIgnal Timing.
Consumer Transport Interfaces
It is important to note that no loops or parallel connections are allowed within the network. Also, there are no reserved connectors—any connector may be used to add a new device to the network. Since 1394-1995 mandates a maximum of 16 cable “hops” between any two nodes, a maximum of 17 peripherals can be included in a network if only two-port peripherals are used. Later specifications implement a “ping” packet to measure the round-trip delay to any node, removing the 16 “hop” limitation. For 1394-1995 and 1394A-2000, a 4- or 6-pin connector is used. The 6-pin connector can provide power to peripherals. For 1394B-2002, the 9-pin Beta and Bilingual connector includes power, two extra pins for signal integrity and one pin for reserved for future use.
Figure 6.54 illustrates the 1394-1995 and 1394A-2000 data and strobe timing. The strobe signal changes state on every bit period for which the data signal does not. Therefore, by exclusive-ORing the data and strobe signals, the clock is recovered. Physical Layer The typical hardware topology of a 1394 network consists of a physical layer (PHY) and link layer (LINK), as shown in Figure 6.55. The 1394-1995 standard also defined two software layers, the transaction layer and the bus management layer, parts of which may be implemented in hardware. The PHY transforms the point-to-point network into a logical physical bus. Each node is also essentially a data repeater since data is
BUSY#
1394 LINK LAYER CONTROL (LLC)
INTREQ# ARXD#
1394 PHYSICAL LAYER INTERFACE
RECEIVED DATA DECODER AND RETIMER
TPA1, TPA1# TPB1, TPB1# TPA2, TPA2# TPB2, TPB2# TPA3, TPA3# TPB3, TPB3#
CABLE PORT 1
CABLE PORT 2
CABLE PORT 3
ARBITRATION AND CONTROL STATE MACHINE LOGIC
TRANSMIT DATA ENCODER
IEC 61883-4 PROTOCOL COPY PROTECT
ASYNCHRONOUS RECEIVE FIFO
BCLK HOST INTERFACE
RESET# D0–D15 A0–A7
ASYNCHRONOUS TRANSMIT FIFO
CS# RD# WR#
1394 LLC CONTROL AND STATUS REGISTERS
ID0–ID7 IWR#
CYCLE MONITOR CYCLE TIMER 1394 PACKET TRANSMIT AND RECEIVE CONTROL LOGIC
187
ISOCHRONOUS RECEIVE FIFO
ISOCHRONOUS TRANSMIT FIFO
ISOCHRONOUS PORT -----------------MPEG 2 TRANSPORT LAYER INTERFACE
CRC LOGIC
Figure 6.55. IEEE 1394 Typical Physical and Link Layer Block Diagrams.
IRXD# IRDY# IDONE# IRESET# ICLK IERROR# IRST# PFTFLAG#
188
Chapter 6: Digital Video Interfaces
reclocked at each node. The PHY also defines the electrical and mechanical connection to the network. Physical signaling circuits and logic responsible for power-up initialization, arbitration, bus-reset sensing, and data signaling are also included.
The delivery latency of asynchronous packets is not guaranteed and depends upon the network traffic. However, the sender may continually retry until an acknowledgment is received. Asynchronous packets are targeted to one node on the network or can be sent to all nodes, but can not be broadcast to a subset of nodes on the bus. The maximum asynchronous packet size is:
Link Layer The LINK provides interfacing between the physical layer and application layer, formatting data into packets for transmission over the network. It supports both asynchronous and isochronous data.
512 * (n / 100) bytes n = network speed in Mbps
Asynchronous Data Asynchronous packets are guaranteed delivery since after an asynchronous packet is received, the receiver transmits an acknowledgment to the sender, as shown in Figure 6.56. However, there is no guaranteed bandwidth. This type of communication is useful for commands, non-real-time data, and error-free transfers.
Isochronous Data Isochronous communications have a guaranteed bandwidth, with up to 80% of the network bandwidth available for isochronous use. Up to 63 independent isochronous channels are available, although the 1394 Open Host Controller Interface (OHCI) currently only supports 4–32 channels. This type of communi-
CYCLE START PACKET
ACK 2
PACKET 2
PACKET 1
ACK 1
ASYNCHRONOUS PACKETS
CHANNEL 3
CHANNEL 2
CHANNEL 1
CYCLE START PACKET
ISOCHRONOUS PACKETS
125 µs
Figure 6.56. IEEE 1394 Isochronous and Asynchronous Packets.
Consumer Transport Interfaces
cation is useful for real-time audio and video transfers since the maximum delivery latency of isochronous packets is calculable and may be targeted to multiple destinations. However, the sender may not retry sending a packet. The maximum isochronous packet size is: 1024 * (n / 100) bytes n = network speed in Mbps
Isochronous operation guarantees a time slice each 125 µs. Since time slots are guaranteed, and isochronous communication takes priority over asynchronous, isochronous bandwidth is assured. Once an isochronous channel is established, the sending device is guaranteed to have the requested amount of bus time for that channel every isochronous cycle. Only one device may send data on a particular channel, but any number of devices may receive data on a channel. A device may use multiple isochronous channels as long as capacity is available. Transaction Layer The transaction layer supports asynchronous write, read, and lock commands. A lock combines a write with a read by producing a round trip routing of data between the sender and receiver, including processing by the receiver. Bus Management Layer The bus management layer control functions of the network at the physical, link, and transaction layers. Digital Transmission Content Protection (DTCP) To prevent unauthorized copying of content, the DTCP system was developed. Although originally designed for 1394, it is
189
applicable to any digital network that supports bi-directional communications, such as USB and Ethernet. Device authentication, content encryption, and renewability (should a device ever be compromised) are supported by DTCP. The Digital Transmission Licensing Administrator (DTLA) licenses the content protection system and distributes cipher keys and device certificates. DTCP outlines four elements of content protection: 1. Copy control information (CCI) 2. Authentication and key exchange (AKE) 3. Content encryption 4. System renewability
Copy Control Information (CCI) CCI allows content owners to specify how their content can be used, such as “copynever,” “copy-one-generation,” “no-more-copies,” and “copy-free.” DTCP is capable of securely communicating copy control information between devices. Two different CCI mechanisms are supported: embedded and encryption mode indicator. Embedded CCI is carried within the content stream. Tampering with the content stream results in incorrect decryption, maintaining the integrity of the embedded CCI. The encryption mode indicator (EMI) provides a secure, yet easily accessible, transmission of CCI by using the two most significant bits of the sync field of the isochronous packet header. Devices can immediately determine the CCI of the content stream without decoding the content. If the two EMI bits are tampered with, the encryption and decryption modes do not match, resulting in incorrect content decryption.
190
Chapter 6: Digital Video Interfaces
Authentication and Key Exchange (AKE) Before sharing content, a device must first verify that the other device is authentic. DTCP includes a choice of two authentication levels: full and restricted. Full authentication can be used with all content protected by the system. Restricted authentication enables the protection of “copy-one-generation” and “no-morecopies” content only. Full authentication Compliant devices are assigned a unique public/private key pair and a device certificate by the DTLA, both stored within the device so as to prevent their disclosure. In addition, devices store other necessary constants and keys. Full authentication uses the public keybased digital signature standard (DSS) and Diffie-Hellman (DH) key exchange algorithms. DSS is a method for digitally signing and verifying the signatures of digital documents to verify the integrity of the data. DH key exchange is used to establish control-channel symmetric cipher keys, which allows two or more devices to generate a shared key. Initially, the receiver sends a request to the source to exchange device certificates and random challenges. Then, each device calculates a DH key exchange first-phase value. The devices then exchange signed messages that contain the following elements: 1. The other device’s random challenge 2. The DH key-exchange first-phase value 3. The renewability message version number of the newest system renewability message (SRM) stored by the device
The devices check the message signatures using the other device’s public key to verify that the message has not been tampered with
and also verify the integrity of the other device’s certificate. Each device also examines the certificate revocation list (CRL) embedded in its system renewability message (SRM) to verify that the other device’s certificate has not been revoked due to its security having been compromised. If no errors have occurred, the two devices have successfully authenticated each other and established an authorization key. Restricted authentication Restricted authentication may be used between sources and receivers for the exchange of “copy-one-generation” and “nomore-copies” contents. It relies on the use of a shared secret to respond to a random challenge. The source initiates a request to the receiver, requests its device ID, and sends a random challenge. After receiving the challenge back from the source, the receiver computes a response and sends it to the source. The source compares this response with similar information generated by the source using its service key and the ID of the receiver. If the comparison matches its own calculation, the receiver has been verified and authenticated. The source and receiver then each calculate an authorization key. Content Encryption To ensure interoperability, all compliant devices must support the 56-bit M6 baseline cipher. Additional content protection may be supported by using additional, optional ciphers. System Renewability Devices that support full authentication can receive and process SRMs that are created by the DTLA and distributed with content. Sys-
Consumer Transport Interfaces
tem renewability is used to ensure the longterm system integrity by revoking the device IDs of compromised devices. SRMs can be updated from other compliant devices that have a newer list, from media with prerecorded content, or via compliant devices with external communication capability (Internet, phone, cable, network, and so on). Example Operation For this example, the source has been instructed to transmit a copy protected system stream of content. The source initiates the transmission of content marked with the copy protection status: “copy-one-generation,” “copy-never,” “nomore-copies,” or “copy-free.” Upon receiving the content stream, the receiver determines the copy protection status. If marked “copy never,” the receiver requests that the source initiate full authentication. If the content is marked “copy once” or “no more copies,” the receiver will request full authentication if supported, or restricted authentication if it isn’t. When the source receives the authentication request, it proceeds with the requested type of authentication. If full authentication is requested but the source can only support restricted authentication, then restricted authentication is used. Once the devices have completed the authentication procedure, a content-channel encryption key (content key) is exchanged between them. This key is used to encrypt the content at the source device and decrypt the content at the receiver. 1394 Open Host Controller Interface (OHCI) The 1394 Open Host Controller Interface (OHCI) specification is an implementation of
191
the 1394 link layer, with additional features to support the transaction and bus management layers. It provides a standardized way of interacting with the 1394 network. Home AV Interoperability (HAVi) Home AV Interoperability (HAVi) is another layer of protocols for 1394. HAVi is directed at making 1394 devices plug-and-play interoperable in a 1394 network whether or not a PC host is present. Serial Bus Protocol (SBP-2) The ANSI Serial Bus Protocol 2 (SBP-2) defines standard way of delivering command and status packets over 1394 for devices such DVD players, printers, scanners, hard drives, and other devices. IEC 61883 Specifications Certain types of isochronous signals, such as MPEG-2 and the IEC 61834, SMPTE 314M and ITU-R BT.1618 digital video (DV) standards, use specific data transport protocols and formats. When this data is sent isochronously over a 1394 network, special packetization techniques are used. The IEC 61883 series of specifications define the details for transferring various application-specific data over 1394: IEC 61883-1 = General specification IEC 61883-2 = SD-DVCR data transmission 25 Mbps continuous bit rate IEC 61883-3 = HD-DVCR data transmission IEC 61883-4 = MPEG-2 TS data transmission bit rate bursts up to 44 Mbps IEC 61883-5 = SDL-DVCR data transmission IEC 61883-6 = Audio and music data transmission IEC 61883-7 = Transmission of ITU-R BO.1294 System B
192
Chapter 6: Digital Video Interfaces
IEC 61883-1 IEC 61883-1 defines the general structure for transferring digital audio and video data over 1394. It describes the general packet format, data flow management, and connection management for digital audio and video data, and also the general transmission rules for control commands. A common isochronous packet (CIP) header is placed at the beginning of the data field of isochronous data packets, as shown in Figure 6.57. It specifies the source node, data block size, data block count, time stamp, type of real-time data contained in the data field, etc. A connection management procedure (CMP) is also defined for making isochronous connections between devices. In addition, a functional control protocol (FCP) is defined for exchanging control commands over 1394 using asynchronous data.
NORMAL ISOCHRONOUS PACKET
IEC 61883-2 IEC 61883-2 and SMPTE 396M define the CIP header, data packet format, and transmission timing for IEC 61834, SMPTE 314M and ITU-R BT.1618 digital video (DV) standards over 1394. Active resolutions of 720 × 480 (at 29.97 frames per second) and 720 × 576 (at 25 frames per second) are supported. DV data packets are 488 bytes long, made up of 8 bytes of CIP header and 480 bytes of DV data, as shown in Figure 6.57. Figure 6.58 illustrates the frame data structure. Each of the 720 × 480 4:1:1 YCbCr frames are compressed to 103,950 bytes, resulting in a 4.9:1 compression ratio. Including overhead and audio increases the amount of data to 120,000 bytes.
61883 - 2 ISOCHRONOUS PACKET
PACKET HEADER HEADER CRC
CIP HEADER 0 CIP HEADER 1
ISOCHRONOUS PACKET PAYLOAD
DATA PAYLOAD (480 BYTES)
DATA CRC 32 BITS
32 BITS
Figure 6.57. 61883-2 Isochronous Packet Formatting.
Consumer Transport Interfaces
193
1 FRAME IN 1.001 / 30 SECOND (10 DIF SEQUENCES)
DIFS0
DIFS1
DIFS2
DIFS3
DIFS4
DIFS5
DIFS6
DIFS7
DIFS8
DIFS9
1 DIF SEQUENCE IN 1.001 / 300 SECOND (150 DIF BLOCKS)
HEADER (1 DIF)
SUBCODE (2 DIF)
VAUX (3 DIF)
135 VIDEO AND 9 AUDIO DIF BLOCKS
150 DIF BLOCKS IN 1.001 / 30 SECOND
DIF0
DIF1
DIF2
DIF3
DIF4
DIF5
DIF6
DIF148
DIF149
1 DIF BLOCK IN 1.001 / 45000 SECOND
ID (3 BYTES)
HEADER (1 BYTE)
Y0 (14 BYTES)
DC0
AC
DATA (76 BYTES)
Y1 (14 BYTES)
DC1
AC
Y2 (14 BYTES)
DC2
AC
Y3 (14 BYTES)
DC3
AC
CR (10 BYTES)
DC4
AC
CB (10 BYTES)
DC5
AC
COMPRESSED MACROBLOCK
Figure 6.58. IEC 61834, SMPTE 314M and ITU-R BT.1618 Packet Formatting for 720 × 480 Systems (4:1:1 YCbCr).
194
Chapter 6: Digital Video Interfaces
The compressed 720 × 480 frame is divided into 10 DIF (data in frame) sequences. Each DIF sequence contains 150 DIF blocks of 80 bytes each, used as follows: 135 DIF blocks for video 9 DIF blocks for audio 6 DIF blocks used for Header, Subcode, and Video Auxiliary (VAUX) information
Figure 6.59 illustrates the DIF sequence structure in detail. The audio DIF blocks contain both audio data and audio auxiliary data (AAUX). IEC 61834 supports four 32-kHz, 12bit nonlinear audio signals or two 48-, 44.1-, or 32-kHz, 16-bit audio signals. SMPTE 314M and ITU-R BT.1618 at 25 Mbps support two 48-kHz 16-bit audio signals, while the 50 Mbps version supports four. Video auxiliary data (VAUX) DIF blocks include recording date and time, lens aperture, shutter speed, color balance, and other camera setting data. The subcode DIF blocks store a variety of information, the most important of which is timecode. Each video DIF block contains 80 bytes of compressed macroblock data: 3 bytes for DIF block ID information 1 byte for the header that includes the quantization number (QNO) and block status (STA) 14 bytes each for Y0, Y1, Y2, and Y3
well, you have a frame buffer with a 120,000 byte compressed DV frame in it. 720 × 576 frames may use either the 4:2:0 YCbCr format (IEC 61834) or the 4:1:1 YCbCr format (SMPTE 314M and ITU-R BT.1618), and require 12 DIF sequences. Each 720 × 576 frame is compressed to 124,740 bytes. Including overhead and audio increases the amount of data to 144,000 bytes, requiring 300 packets to transfer. Note that the organization of data transferred over 1394 differs from the actual DV recording format since error correction is not required for digital transmission. In addition, although the video blocks are numbered in sequence in Figure 6.59, the sequence does not correspond to the left-to-right, top-to-bottom transmission of blocks of video data. Compressed macroblocks are shuffled to minimize the effect of errors and aid in error concealment. Audio data also is shuffled. Data is transmitted in the same shuffled order as recorded. To illustrate the video data shuffling, DV video frames are organized as 50 super blocks, with each super block being composed of 27 compressed macroblocks, as shown in Figure 6.60. A group of 5 super blocks (one from each super block column) make up one DIF sequence. Table 6.42 illustrates the transmission order of the DIF blocks. Additional information on the DV data structure is available in Chapter 11.
10 bytes each for Cb and Cr
As the 488-byte packets come across the 1394 network, the start of a video frame is determined. Once the start of a frame is detected, 250 valid packets of data are collected to have a complete DV frame; each packet contains 6 DIF blocks of data. Every 15th packet is a null packet and should be discarded. Once 250 valid packets of data are in the buffer, discard the CIP headers. If all went
IEC 61883-4 IEC 61883-4 defines the CIP header, data packet format, and transmission timing for MPEG-2 transport streams over 1394. It is most efficient to carry an integer number of 192 bytes (188 bytes of MPEG-2 data plus 4 bytes of time stamp) per isochronous packet, as shown in Figure 6.61. However, MPEG data rates are rarely integer multiples of the isochronous data rate. Thus, it is more
195
Consumer Transport Interfaces
H = HEADER SECTION H
SC0
SC1
VA0
VA1
VA2
0
1
2
3
4
5
SC0, SC1 = SUBCODE SECTION VA0, VA1, VA2 = VAUX SECTION A0–A8 = AUDIO SECTION V0–V134 = VIDEO SECTION
A0
V0
V1
V2
V3
V4
V5
V6
V7
V8
V9
V10
V11
V12
V13
V14
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
A1
V15
V16
V17
V18
V19
V20
V21
V22
V23
V24
V25
V26
V27
V28
V29
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
A8
V120
V121
V122
V123
V124
V125
V126
V127
V128
V129
V130
V131
V132
V133
V134
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
Figure 6.59. IEC 61834, SMPTE 314M and ITU-R BT.1618 DIF Sequence Detail (25 Mbps).
196
Chapter 6: Digital Video Interfaces
720 SAMPLES SUPERBLOCK
480 LINES
0
1
2
3
4
0
S0,0
S0,1
S0,2
S0,3
S0,4
1
S1,0
S1,1
S1,2
S1,3
S1,4
2
S2,0
S2,1
S2,2
S2,3
S2,4
3
S3,0
S3,1
S3,2
S3,3
S3,4
4
S4,0
S4,1
S4,2
S4,3
S4,4
5
S5,0
S5,1
S5,2
S5,3
S5,4
6
S6,0
S6,1
S6,2
S6,3
S6,4
7
S7,0
S7,1
S7,2
S7,3
S7,4
8
S8,0
S8,1
S8,2
S8,3
S8,4
9
S9,0
S9,1
S9,2
S9,3
S9,4
0
11 12 23 24
8
9
20 21
0
11 12 23 24
8
9
20 21
0
11 12 23
1
10 13 22 25
7
10 19 22
1
10 13 22 25
7
10 19 22
1
10 13 22
2
9
14 21 26
6
11 18 23
2
9
14 21 26
6
11 18 23
2
9
14 21
3
8
15 20
0
5
12 17 24
3
8
15 20
0
5
12 17 24
3
8
15 20
4
7
16 19
1
4
13 16 25
4
7
16 19
1
4
13 16 25
4
7
16 19
5
6
17 18
2
3
14 15 26
5
6
17 18
2
3
14 15 26
5
6
17 18
MACROBLOCK
24 25 26
Figure 6.60. Relationship Between Super Blocks and Macroblocks (720 × 480, 4:1:1 YCbCr).
Consumer Transport Interfaces
DIF Sequence Number
0
Compressed Macroblock Superblock Number
Macroblock Number
0
2, 2
0
1
6, 1
0
0
1, 2
0
2
8, 3
0
1
5, 1
0
3
0, 0
0
2
7, 3
0
4
4, 4
0
3
n–1, 0
0
4
3, 4
0
: 133
1
Compressed Macroblock
Video DIF Block Number
DIF Sequence Number
Video DIF Block Number
Superblock Number
Macroblock Number
:
n–1
:
0, 0
26
134
4, 4
26
133
n–1, 0
26
0
3, 2
0
134
3, 4
26
1
7, 1
0
2
9, 3
0
3
1, 0
0
4
5, 4
0
: 133
1, 0
26
134
5, 4
26
Notes: 1. n = 10 for 480-line systems, n = 12 for 576-line systems.
Table 6.42. Video DIF Blocks and Compressed Macroblocks for 25 Mbps.
197
198
Chapter 6: Digital Video Interfaces
efficient to divide the MPEG packets into smaller components of 24 bytes each to maximize available bandwidth. The transmitter then uses an integer number of data blocks (restricted multiples of 0, 1, 2, 4, or 8) placing them in an isochronous packet and adding the 8-byte CIP header. 50 Mbps DV Like the 25 Mbps DV format, the 50 Mbps DV format supports 720 × 480i30 and 720 × 576i25 sources. However, the 50 Mbps DV format uses 4:2:2 YCbCr rather than 4:1:1 YCbCr. As previously discussed, the source packet size for the 25 Mbps DV streams is 480 bytes (consisting of 6 DIF blocks). The 250 packets (300 packets for 576i25 systems) of 480-byte data are transferred over a 25 Mbps channel. The source packet size for the 50 Mbps DV streams is 960 bytes (consisting of 12 DIF blocks). The first 125 packets (150 packets for 576i25 systems) of 960-byte data are sent over one 25 Mbps channel and the next 125 packets (150 packets for 576i25 systems) of 960-byte data are sent over a second 25 Mbps channel.
NORMAL ISOCHRONOUS PACKET
100 Mbps DV 100 Mbps DV streams support 1920 × 1080i30, 1920 × 1080i25 and 1280 × 720p60 sources. 1920 × 1080i30 sources are horizontally scaled to 1280 × 1080i30. 1920 × 1080i25 sources are horizontally scaled to 1440 × 1080i25. 1280 × 720p60 sources are horizontally scaled to 960 × 720p60. The 4:2:2 YCbCr format is used. The source packet size for the 100 Mbps DV streams is 1920 bytes (consisting of 24 DIF blocks). The first 63 packets (75 packets for 1080i25 systems) of 1920-byte data are sent over one 25 Mbps channel, the next 62 packets (75 packets for 1080i25 systems) of 1920-byte data are sent over a second 25 Mbps channel, the next 63 packets (75 packets for 1080i25 systems) of 1920-byte data are sent over a third 25 Mbps channel and the last 62 packets (75 packets for 1080i25 systems) of 1920-byte data are sent over a fourth 25 Mbps channel.
61883 - 4 ISOCHRONOUS PACKET
PACKET HEADER HEADER CRC
CIP HEADER 0 CIP HEADER 1
ISOCHRONOUS PACKET PAYLOAD
DATA PAYLOAD (192 BYTES)
DATA CRC 32 BITS
32 BITS
Figure 6.61. 61883-4 Isochronous Packet Formatting.
References
Digital Camera Specification The 1394 Trade Association has written a specification for 1394-based digital video cameras. This was done to avoid the silicon and software cost of implementing the full IEC 61883 specification. Seven resolutions are defined, with a wide range of format support: 160 × 120
320 × 240 640 × 480
800 × 600
1024 × 768
1280 × 960
1600 × 1200
4:4:4 YCbCr 4:2:2 YCbCr 4:1:1, 4:2:2 YCbCr, 24-bit RGB 4:2:2 YCbCr, 24-bit RGB 4:2:2 YCbCr, 24-bit RGB 4:2:2 YCbCr, 24-bit RGB 4:2:2 YCbCr, 24-bit RGB
Supported frame rates are 1.875, 3.75, 7.5, 15, 30, and 60 frames per second. Isochronous packets are used to transfer the uncompressed digital video data over the 1394 network.
References 1. 1394-based Digital Camera Specification, Version 1.20, July 23, 1998. 2. Digital Transmission Content Protection Specification, Volume 1 (Informational Version), July 25, 2000. 3. Digital Visual Interface (DVI), April 2, 1999. 4. EBU Tech. 3267-E, 1992, EBU Interfaces for 625-Line Digital Video Signals at the 4:2:2 Level of CCIR Recommendation 601, European Broadcasting Union, June, 1991. 5. IEC 61883–1, 2003, Consumer Audio/Video Equipment—Digital Interface—Part 1: General.
199
6. IEC 61883–2, 1998, Consumer Audio/Video Equipment—Digital Interface—Part 2: SDDVCR Data Transmission. 7. IEC 61883–3, 1998, Consumer Audio/Video Equipment—Digital Interface—Part 3: HD-DVCR Data Transmission. 8. IEC 61883–4, 1998, Consumer Audio/Video Equipment—Digital Interface—Part 4: MPEG-2 TS Data Transmission. 9. IEC 61883–5, 1998, Consumer Audio/Video Equipment—Digital Interface—Part 5: SDL-DVCR Data Transmission. 10. ITU-R BT.656–4, 1998, Interfaces for Digital Component Video Signals in 525-Line and 625-Line Television Systems Operating at the 4:2:2 Level of Recommendation ITU-R BT.601. 11. ITU-R BT.799–3, 1998, Interfaces For Digital Component Video Signals in 525Line and 625-Line Television Systems Operating at the 4:4:4 Level of Recommendation ITU-R BT.601 (Part A). 12. ITU-R BT.1302, 1997, Interfaces for Digital Component Video Signals in 525-Line and 625-Line Television Systems Operating at the 4:2:2 Level of ITU-R BT.601. 13. ITU-R BT.1303, 1997, Interfaces For Digital Component Video Signals in 525-Line and 625-Line Television Systems Operating at the 4:4:4 Level of Recommendation ITU-R BT.601 (Part B). 14. ITU-R BT.1304, 1997, Checksum for Error Detection and Status Information in Interfaces Conforming to ITU-R BT.656 and ITU-R BT.799. 15. ITU-R BT.1305, 1997, Digital Audio and Auxiliary Data as Ancillary Data Signals in Interfaces Conforming to ITU-R BT.656 and ITU-R BT.799. 16. ITU-R BT.1362, 1998, Interfaces For Digital Component Video Signals in 525-Line and 625-Line Progressive Scan Television Systems.
200
Chapter 6: Digital Video Interfaces
17. ITU-R BT.1364, 1998, Format of Ancillary Data Signals Carried in Digital Component Studio Interfaces. 18. ITU-R BT.1365, 1998, 24-Bit Digital Audio Format as Ancillary Data Signals in HDTV Serial Interfaces. 19. ITU-R BT.1366, 1998, Transmission of Time Code and Control Code in the Ancillary Data Space of a Digital Television Stream According to ITU-R BT.656, ITU-R BT.799, and ITU-R BT.1120. 20. ITU-R BT.1381–1, 2001, Serial Digital Interface-based Transport Interface for Compressed Television Signals In Networked Television Production Based On Recommendations ITU-R BT.656 and ITUR BT.1302. 21. ITU-R BT.1577, 2002, Serial digital interface-based transport interface for compressed television signals in networked television production based on Recommendation ITU-R BT.1120. 22. ITU-R BT.1616, 2003, Data stream format for the exchange of DV-based audio, data and compressed video over interfaces complying with Recommendation ITU-R BT.1381. 23. ITU-R BT.1617, 2003, Format for transmission of DV compressed video, audio and data over interfaces complying with Recommendation ITU-R BT.1381. 24. ITU-R BT.1618, 2003, Data structure for DV-based audio, data and compressed video at data rates of 25 and 50 Mbit/s. 25. ITU-R BT.1619, 2003, Vertical ancillary data mapping for serial digital interface. 26. ITU-R BT.1620, 2003, Data structure for DV-based audio, data and compressed video at a data rate of 100 Mbit/s. 27. Kikuchi, Hidekazu et. al., A 1-bit Serial Interface Chip Set for Full-Color XGA Pictures, Society for Information Display, 1999.
28. Kikuchi, Hidekazu et. al., Gigabit Video Interface: A Fully Serialized Data Transmission System for Digital Moving Pictures, International Conference on Consumer Electronics, 1998. 29. Open LVDS Display Interface (OpenLDI) Specification, v0.95, May 13, 1999. 30. SMPTE 125M–1995, Television—Component Video Signal 4:2:2—Bit-Parallel Digital Interface. 31. SMPTE 240M–1999, Television—Signal Parameters—1125-Line High-Definition Production Systems. 32. SMPTE 244M–2003, Television—System M/NTSC Composite Video Signals—BitParallel Digital Interface. 33. SMPTE 259M–1997, Television—10-Bit 4:2:2 Component and 4FSC Composite Digital Signals—Serial Digital Interface. 34. SMPTE 260M–1999, Television—1125/60 High-Definition Production System—Digital Representation and Bit-Parallel Interface. 35. SMPTE 266M–2002, Television—4:2:2 Digital Component Systems—Digital Vertical Interval Time Code. 36. SMPTE 267M–1995, Television—Bit-Parallel Digital Interface—Component Video Signal 4:2:2 16 × 9 Aspect Ratio. 37. SMPTE 272M–1994, Television—Formatting AES/EBU Audio and Auxiliary Data into Digital Video Ancillary Data Space. 38. SMPTE 274M–2003, Television—1920 × 1080 Image Sample Structure, Digital Representation and Digital Timing Reference Sequences for Multiple Picture Rates. 39. SMPTE 291M–1998, Television—Ancillary Data Packet and Space Formatting. 40. SMPTE 292M–1998, Television—Bit-Serial Digital Interface for High-Definition Television Systems. 41. SMPTE 293M–2003, Television—720 × 483 Active Line at 59.94-Hz Progressive Scan Production—Digital Representation.
References
42. SMPTE 294M–2001, Television—720 × 483 Active Line at 59.94 Hz Progressive Scan Production—Bit-Serial Interfaces. 43. SMPTE 296M–2001, Television—1280 × 720 Progressive Image Sample Structure, Analog and Digital Representation and Analog Interface. 44. SMPTE 305.2M–2000, Television—Serial Data Transport Interface (SDTI). 45. SMPTE 314M–1999, Television—Data Structure for DV-Based Audio, Data and Compressed Video—25 and 50 Mb/s. 46. SMPTE 326M–2000, Television—SDTI Content Package Format (SDTI-CP). 47. SMPTE 334M–2000, Television—Vertical Ancillary Data Mapping for Bit-Serial Interface. 48. SMPTE 344M–2000, Television—540 Mbps Serial Digital Interface. 49. SMPTE 348M–2000, Television—High Data-Rate Serial Data Transport Interface (HD-SDTI). 50. SMPTE 370M–2002, Television—Data Structure for DV-Based Audio, Data and Compressed Video at 100 Mb/s 1080/60i, 1080/50i, 720-60p. 51. SMPTE 372M–2002, Television—Dual Link 292M Interface for 1920 × 1080 Picture Raster. 52. SMPTE 396M–2003, Television—Packet Format and Transmission Timing of DVBased Data Streams over IEEE 1394.
201
53. SMPTE RP-165–1994, Error Detection Checkwords and Status Flags for Use in BitSerial Digital Interfaces for Television. 54. SMPTE RP-174–1993, Bit-Parallel Digital Interface for 4:4:4:4 Component Video Signal (Single Link). 55. SMPTE RP-175–1997, Digital Interface for 4:4:4:4 Component Video Signal (Dual Link). 56. SMPTE RP-168–2002, Definition of Vertical Interval Switching Point for Synchronous Video Switching. 57. SMPTE RP-188–1999, Transmission of Time Code and Control Code in the Ancillary Data Space of a Digital Television Data Stream. 58. SMPTE RP-208–2002, Transport of VBI Packet Data in Ancillary Data Packets. 59. Teener, Michael D. Johas, IEEE 13941995 High Performance Serial Bus, 1394 Developer’s Conference, 1997. 60. VESA DFP 1.0: Digital Flat Panel (DFP) Standard. 61. VESA Video Interface Port (VIP), Version 2, October 21, 1998. 62. VMI Specification, v1.4, January 30, 1996. 63. Wickelgren, Ingrid J., The Facts about FireWire, IEEE Spectrum, April 1997.
202
Chapter 7: Digital Video Processing
Chapter 7: Digital Video Processing Chapter 7
Digital Video Processing In addition to encoding and decoding MPEG, NTSC/PAL and many other types of video, a typical system usually requires considerable additional video processing. Since many consumer displays, and most computer displays, are progressive (noninterlaced), interlaced video must be converted to progressive (“deinterlaced”). Progressive video must be converted to interlaced to drive a conventional analog VCR or TV, requiring noninterlaced-to-interlaced conversion. Many computer displays have a vertical refresh rate of about 75 Hz, whereas consumer video has a vertical refresh rate of 25 or 29.97 (30/1.001) frames per second. For DVD and DTV, source material may only be 24 frames per second. Thus, some form of frame rate conversion must be done.
202
Another not-so-subtle problem includes video scaling. SDTV and HDTV support multiple resolutions, yet the display may be a single, fixed resolution. Alpha mixing and chroma keying are used to mix multiple video signals or video with computer-generated text and graphics. Alpha mixing ensures a smooth crossover between sources, allows subpixel positioning of text, and limits source transition bandwidths to simplify eventual encoding to composite video signals. Since no source is perfect, even digital sources, user controls for adjustable brightness, contrast, saturation, and hue are always desirable.
Rounding Considerations
positive numbers should be made less positive and negative numbers should be made less negative.
Rounding Considerations When two 8-bit values are multiplied together, a 16-bit result is generated. At some point, a result must be rounded to some lower precision (for example, 16 bits to 8 bits or 32 bits to 16 bits) in order to realize a cost-effective hardware implementation. There are several rounding techniques: truncation, conventional rounding, error feedback rounding, and dynamic rounding.
Error Feedback Rounding Error feedback rounding follows the principle of “never throw anything away.” This is accomplished by storing the residue of a truncation and adding it to the next video sample. This approach substitutes less visible noiselike quantizing errors in place of contouring effects caused by simple truncation. An example of an error feedback rounding implementation is shown in Figure 7.1. In this example, 16 bits are reduced to 8 bits using error feedback.
Truncation Truncation drops any fractional data during each rounding operation. As a result, after only a few operations, a significant error may be introduced. This may result in contours being visible in areas of solid colors.
Dynamic Rounding This technique (a licensable Quantel patent) dithers the LSB according to the weighting of the discarded fractional bits. The original data word is divided into two parts, one representing the resolution of the final output word and one dealing with the remaining fractional data. The fractional data is compared to the output of a random number generator equal in resolution to the fractional data. The output of the comparator is a 1-bit random pattern weighted by the value of the fractional
Conventional Rounding Conventional rounding uses the fractional data bits to determine whether to round up or round down. If the fractional data is 0.5 or greater, rounding up should be performed— positive numbers should be made more positive and negative numbers should be made more negative. If the fractional data is less than 0.5, rounding down should be performed—
16–BIT DATA IN
16
16
203
8 (MSB)
+
8–BIT DATA OUT
8 (LSB) 8 (LSB) 8 MSB = 0 REGISTER
Figure 7.1. Error Feedback Rounding.
204
Chapter 7: Digital Video Processing
data, and serves as a carry-in to the adder. In all instances, only one LSB of the output word is changed, in a random fashion. An example of a dynamic rounding implementation is shown in Figure 7.2.
1 – 0.11554975 – 0.20793764 0 1.01863972 0.11461795 0 0.07504945 1.02532707
Note that before processing, the 8-bit DC offset (16 for Y and 128 for CbCr) must be removed, then added back in after processing.
SDTV - HDTV YCbCr Transforms SDTV and HDTV applications have different colorimetric characteristics, as discussed in Chapter 3. Thus, when SDTV (HDTV) data is displayed on a HDTV (SDTV) display, the YCbCr data should be processed to compensate for the different colorimetric characteristics.
HDTV to SDTV
A 3 × 3 matrix can be used to convert from Y709CbCr (HDTV) to Y601CbCr (SDTV):
1 0.09931166 0.19169955 0 0.98985381 – 0.11065251
SDTV to HDTV
A 3 × 3 matrix can be used to convert from Y601CbCr (SDTV) to Y709CbCr (HDTV):
16–BIT DATA IN
16
0 – 0.07245296 0.98339782
Note that before processing, the 8-bit DC offset (16 for Y and 128 for CbCr) must be removed, then added back in after processing.
8
8 (MSB)
+ 8 (LSB)
CARRY IN
A>B A
PSEUDO RANDOM BINARY SEQUENCE GENERATOR
COMPARATOR
8 B
Figure 7.2. Dynamic Rounding.
8–BIT DATA OUT
4:4:4 to 4:2:2 YCbCr Conversion
ances to avoid a buildup of visual artifacts. Departure from flat amplitude and group delay response due to filtering is amplified through successive stages. For example, if filters exhibiting –1 dB at 1 MHz and –3 dB at 1.3 MHz were employed, the overall response would be –8 dB (at 1 MHz) and –24 dB (at 1.3 MHz) after four conversion stages (assuming two filters per stage). Although the sharp cut-off results in ringing on Y edges, the visual effect should be minimal provided that group-delay performance is adequate. When cascading multiple filtering operations, the passband flatness and groupdelay characteristics are very important. The passband tolerances, coupled with the sharp cut-off, make the template very difficult (some say impossible) to match. As a result, there is usually a temptation to relax passband accuracy, but the best approach is to reduce the rate of cut-off and keep the passband as flat as possible.
4:4:4 to 4:2:2 YCbCr Conversion Converting 4:4:4 YCbCr to 4:2:2 YCbCr (Figure 7.3) is a common function in digital video. 4:2:2 YCbCr is the basis for many digital video interfaces, and requires fewer connections to implement than 4:4:4. Saturation logic should be included in the Y, Cb, and Cr data paths to limit the 8-bit range to 1–254. The 16 and 128 values shown in Figure 7.3 are used to generate the proper levels during blanking intervals.
Y Filtering A template for the Y lowpass filter is shown in Figure 7.4 and Table 7.1. Because there may be many cascaded conversions (up to 10 were envisioned), the filters were designed to adhere to very tight toler-
8-BIT 4:2:2 YCBCR
16-BIT 4:2:2 YCBCR
24-BIT 4:4:4 YCBCR
8 OPTIONAL LPF
Y
8 Y
MUX 16
8 8
MUX
OPTIONAL LPF
CR
8 128
MUX
CBCR
8 CB
205
OPTIONAL LPF
Figure 7.3. 4:4:4 to 4:2:2 YCbCr Conversion.
YCBCR
206
Chapter 7: Digital Video Processing
ATTENUATION (DB) 50 DB
50 40 DB 40 30 20
12 DB
10 0
0.40 FS
0.50 FS
0.60 FS
0.73 FS
FREQUENCY (MHZ)
Figure 7.4. Y Filter Template. Fs = Y 1× sample rate.
Frequency Range
Typical SDTV Tolerances
0 to 0.40Fs
±0.01 dB increasing to ±0.05 dB
Passband Ripple Tolerance
Passband Group Delay Tolerance 0 to 0.27Fs 0.27Fs to 0.40Fs
0 increasing to ±1.35 ns
±1.35 ns increasing to ±2 ns
Typical HDTV Tolerances ±0.05 dB ±0.075T
±0.110T
Table 7.1. Y Filter Ripple and Group Delay Tolerances. Fs = Y 1× sample rate. T = 1 / Fs.
4:4:4 to 4:2:2 YCbCr Conversion
ATTENUATION (DB) 60 55 DB 50 40 30 20 10 6 DB
0
0.30 FS 0.20 FS 0.25 FS
0.47 FS
FREQUENCY (MHZ)
Figure 7.5. Cb and Cr Filter Template for Digital Filter for Sample Rate Conversion from 4:4:4 to 4:2:2. Fs = Y 1× sample rate.
Frequency Range
Typical SDTV Tolerances Passband Ripple Tolerance
0 to 0.20Fs
0 dB increasing to ±0.05 dB
Typical HDTV Tolerances ±0.05 dB
Passband Group Delay Tolerance 0 to 0.20Fs
delay distortion is zero by design
Table 7.2. CbCr Filter Ripple and Group Delay Tolerances. Fs = Y 1× sample rate. T = 1 / Fs.
207
208
Chapter 7: Digital Video Processing
CbCr Filtering Cb and Cr are lowpass filtered and decimated. In a standard design, the lowpass and decimation filters may be combined into a single filter, and a single filter may be used for both Cb and Cr by multiplexing. As with Y filtering, the Cb and Cr lowpass filtering requires a sharp cut-off to prevent repeated conversions from producing a cumulative resolution loss. However, due to the low cut-off frequency, the sharp cut-off produces ringing that is more noticeable than for Y. A template for the Cb and Cr filters is shown in Figure 7.5 and Table 7.2. Since aliasing is less noticeable in color difference signals, the attenuation at half the sampling frequency is only 6 dB. There is an advantage in using a skew-symmetric response passing through the –6 dB point at half the sampling frequency—this makes alternate coefficients in the digital filter zero, almost halving the number of taps, and also allows using a single digital filter for both the Cb and Cr signals. Use of a transversal digital filter has the advantage of providing perfect linear phase response, eliminating the need for group-delay correction. As with the Y filter, the passband flatness and group-delay characteristics are very important, and the best approach again is to reduce the rate of cut-off and keep the passband as flat as possible.
Display Enhancement Brightness, Contrast, Saturation (Color) and Hue (Tint) Working in the YCbCr color space simplifies the implementation of brightness, contrast, saturation and hue controls, as shown in Fig-
ure 7.6. Also illustrated are multiplexers to allow the output of black screen, blue screen and color bars. The design should ensure that no overflow or underflow wrap-around errors occur, effectively saturating results to the 0 and 255 values. Y Processing 16 is subtracted from the Y data to position the black level at zero. This removes the DC offset so adjusting the contrast does not vary the black level. Since the Y input data may have values below 16, negative Y values should be supported at this point. The contrast (or “picture” or “white level”) control is implemented by multiplying the YCbCr data by a constant. If Cb and Cr are not adjusted, a color shift will result whenever the contrast is changed. A typical 8-bit contrast adjustment range is 0–1.992×. The brightness (or “black level”) control is implemented by adding or subtracting from the Y data. Brightness is done after the contrast to avoid introducing a varying DC offset due to adjusting the contrast. A typical 6-bit brightness adjustment range is –32 to +31. Finally, 16 is added to position the black level at 16. CbCr Processing 128 is subtracted from Cb and Cr to position the range about zero. The hue (or “tint”) control is implemented by mixing the Cb and Cr data: Cb´ = Cb cos θ + Cr sin θ Cr´= Cr cos θ – Cb sin θ
where θ is the desired hue angle. A typical 8-bit hue adjustment range is –30° to +30°. The saturation (or “color”) control is implemented by multiplying both Cb and Cr by a constant. A typical 8-bit saturation adjustment
209
Display Enhancement
00 01 10 11 BRIGHTNESS VALUE
CONTRAST VALUE
16
= = = =
BLACK SCREEN BLUE SCREEN COLOR BARS NORMAL VIDEO
16 16 163
8
+
–
COLOR BAR Y
+
Y
8 MUX
Y
+ SATURATION VALUE
HUE VALUE
HUE CONTROL 128
SIN
COS 128
8
+
128
–
105
CR
COLOR BAR CR
+
128 167
8
+
CR
+ –
CB
8 MUX
COLOR BAR CB
–
+
+
Figure 7.6. Hue, Saturation, Contrast, and Brightness Controls.
8 MUX
CB
210
Chapter 7: Digital Video Processing
range is 0–1.992×. In the example shown in Figure 7.6, the contrast and saturation values are multiplied together to reduce the number of multipliers in the CbCr datapath. Finally, 128 is added to both Cb and Cr.
Since this technique artificially increases the high-frequency component of video signals, it should not be used if the video will be compressed, as the compression ratio will be reduced.
Color Transient Improvement
Sharpness
YCbCr transitions should be aligned. However, the Cb and Cr transitions are usually degraded due to the narrow bandwidth of color difference information. By monitoring coincident Y transitions, faster transitions may be synthesized for Cb and Cr. These edges are then aligned with the Y edge, as shown in Figure 7.7. Alternately, Cb and Cr transitions may be differentiated, and the results added to the original Cb and Cr signals. Small amplitudes in the differentiation signals should be suppressed by coring. To eliminate “wrong colors” due to overshoots and undershoots, the enhanced CbCr signals should also be limited to the proper range. In some cases, the Y rise and fall times are also shortened to artificially sharpen the image.
The apparent sharpness of a picture may be increased by increasing the amplitude of high-frequency luminance information. As shown in Figure 7.8, a simple bandpass filter with selectable gain (also called a peaking filter) may be used. The frequency where maximum gain occurs is usually selectable to be either at the color subcarrier frequency or at about 2.6 MHz. A coring circuit is typically used after the filter to reduce low-level noise. Figure 7.9 illustrates a more complex sharpness control circuit. The high-frequency luminance is increased using a variable bandpass filter, with adjustable gain. The coring function (typically ±1 LSB) removes low-level noise. The modified luminance is then added to the original luminance signal. Since this technique artificially increases the high-frequency component of the video signals, it should not be used if the video will be compressed, as the compression ratio will be reduced. In addition to selectable gain, selectable attenuation of high frequencies should also be supported. Many televisions boost high-frequency gain to improve the apparent sharpness of the picture. If this is applied to a compressed (such as MPEG) source, the picture quality can be substantially degraded. Although the sharpness control on the television may be turned down, this affects the picture quality of analog broadcasts. Therefore, many MPEG sources have the option of attenuating high frequencies to negate the sharpness control on the television.
150 NS
Y
CB, CR 800 NS
ENHANCED CB, CR 150 NS
Figure 7.7. Color Transient Improvement.
Display Enhancement
GAIN (DB)
GAIN (DB)
12
12
10
10
8
8
6
6
4
4
2
2 MHZ
0 0
211
1
2
3
4
5
6
MHZ
0
7
0
1
2
(A)
3
4
5
6
7
(B)
Figure 7.8. Simple Adjustable Sharpness Control. (a) NTSC. (b) PAL.
Y IN
VARIABLE BANDPASS FILTER
CORING
WEIGHTING AND ADDING
Y OUT
DELAY
(A) OUT
IN
(B)
Figure 7.9. More Complex Sharpness Control. (a) Typical implementation. (b) Coring function.
212
Chapter 7: Digital Video Processing
Video Mixing and Graphics Overlay Mixing video signals may be as simple as switching between two video sources. This is adequate if the resulting video is to be displayed on a computer monitor. For most other applications, a technique known as alpha mixing should be used. Alpha mixing may also be used to fade to or from a specific color (such as black) or to overlay computer-generated text and graphics onto a video signal. Alpha mixing must be used if the video is to be encoded to composite video. Otherwise, ringing and blurring may appear at the source switching points, such as around the edges of computer-generated text and graphics. This is due to the color information being lowpass filtered within the NTSC/PAL encoder. If the filters have a sharp cut-off, a fast color transition will produce ringing. In addition, the intensity information may be bandwidth-limited to about 4–5 MHz somewhere along the video path, slowing down intensity transitions. Mathematically, with alpha normalized to have values of 0–1, alpha mixing is implemented as: out = (alpha_0)(in_0) + (alpha_1)(in_1) + ...
In this instance, each video source has its own alpha information. The alpha information may not total to one (unity gain). Figure 7.10 shows mixing of two YCbCr video signals, each with its own alpha information. As YCbCr uses an offset binary notation, the offset (16 for Y and 128 for Cb and Cr) is removed prior to mixing the video signals. After mixing, the offset is added back in. Note that two 4:2:2 YCbCr streams may also be processed directly; there is no need to convert
them to 4:4:4 YCbCr, mix, then convert the result back to 4:2:2 YCbCr. When only two video sources are mixed and alpha_0 + alpha_1 = 1 (implementing a crossfader), a single alpha value may be used, mathematically shown as: out = (alpha)(in_0) + (1 – alpha)(in_1)
When alpha = 0, the output is equal to the in_1 video signal; when alpha = 1, the output is equal to the in_0 video signal. When alpha is between 0 and 1, the two video signals are proportionally multiplied, and added together. Expanding and rearranging the previous equation shows how a two-channel mixer may be implemented using a single multiplier: out = (alpha)(in_0 – in_1) + in_1
Fading to and from a specific color is done by setting one of the input sources to a constant color. Figure 7.11 illustrates mixing two YCbCr sources using a single alpha channel. Figures 7.12 and 7.13 illustrate mixing two R´G´B´ video sources (R´G´B´ has a range of 0–255). Figures 7.14 and 7.15 show mixing two digital composite video signals. A common problem in computer graphics systems that use alpha is that the frame buffer may contain preprocessed R´G´B´ or YCbCr data; that is, the R´G´B´ or YCbCr data in the frame buffer has already been multiplied by alpha. Assuming an alpha value of 0.5, nonprocessed R´G´B´A values for white are (255, 255, 255, 128); preprocessed R´G´B´A values for white are (128, 128, 128, 128). Therefore, any mixing circuit that accepts R´G´B´ or YCbCr data from a frame buffer should be able to handle either format. By adjusting the alpha values, slow to fast crossfades are possible, as shown in Figure
Video Mixing and Graphics Overlay
ALPHA_0
16
8
Y_0
–
+
+
8
8
Y_OUT
+
ALPHA_1
16
8
ROUNDING AND LIMITING
16
–
+
Y_1
128
8
ALPHA_0
–
+
CR_0
+
128
8
ROUNDING AND LIMITING
8
8
+
ALPHA_1
CR_OUT
128
–
+
CR_1
128
8
+
ALPHA_0
–
+
CB_0
128
8
ALPHA_1
ROUNDING AND LIMITING
8
8
+
CB_OUT
128
–
+
CB_1
Figure 7.10. Mixing Two YCbCr Video Signals, Each With Its Own Alpha Channel.
213
214
Chapter 7: Digital Video Processing
8 Y_1
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
8 Y_OUT
ALPHA (0–1)
8
–
+
Y_0
8 CR_1
8 CR_OUT
ALPHA (0–1)
8
–
+
CR_0
8 CB_1
8 CB_OUT
ALPHA (0–1)
8
–
+
CB_0
Figure 7.11. Simplified Mixing (Crossfading) of Two YCbCr Video Signals Using a Single Alpha Channel.
Video Mixing and Graphics Overlay
ALPHA_0
8 R_0
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
8 R_OUT
ALPHA_1
8 R_1
ALPHA_0
8 G_0
8 G_OUT
ALPHA_1
8 G_1
ALPHA_0
8 B_0
8 B_OUT
ALPHA_1
8 B_1
Figure 7.12. Mixing Two RGB Video Signals (RGB has a Range of 0–255), Each With Its Own Alpha Channel.
215
216
Chapter 7: Digital Video Processing
8 R_1
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
+
ROUNDING AND LIMITING
8 R_OUT
ALPHA (0–1)
8
–
+
R_0
8 G_1
8 G_OUT
ALPHA (0–1)
8
–
+
G_0
8 B_1
8 B_OUT
ALPHA (0–1)
8
–
+
B_0
Figure 7.13. Simplified Mixing (Crossfading) of Two RGB Video Signals (RGB has a Range of 0–255) Using a Single Alpha Channel.
Video Mixing and Graphics Overlay
BLACK LEVEL
8
ALPHA_0
–
+
ROUNDING AND LMITING
+
SOURCE_0
BLACK LEVEL
8
217
8
8
+
OUT
ALPHA_1 BLACK LEVEL
–
+
SOURCE_1
Figure 7.14. Mixing Two Digital Composite Video Signals, Each With Its Own Alpha Channel.
8
+
SOURCE_1
ROUNDING AND LIMITING
8 OUT
ALPHA (0–1)
8
–
+
SOURCE_0
Figure 7.15. Simplified Mixing (Crossfading) of Two Digital Composite Video Signals Using a Single Alpha Channel.
218
Chapter 7: Digital Video Processing
NORMALIZED ALPHA VALUE
1 50% 0
SAMPLE CLOCK
(A)
NORMALIZED ALPHA VALUE
1 50% 0
SAMPLE CLOCK
(B)
Figure 7.16. Controlling Alpha Values to Implement (a) Fast or (b) Slow Keying. In (a), the effective switching point lies between two samples. In (b), the transition is wider and is aligned at a sample instant.
Luma and Chroma Keying
7.16. Large differences in alpha between samples result in a fast crossfade; smaller differences result in a slow crossfade. If using alpha mixing for special effects, such as wipes, the switching point (where 50% of each video source is used) must be able to be adjusted to an accuracy of less than one sample to ensure smooth movement. By controlling the alpha values, the switching point can be effectively positioned anywhere, as shown in Figure 7.16a. Text can be overlaid onto video by having a character generator control the alpha inputs. By setting one of the input sources to a constant color, the text will assume that color. Note that for those designs that subtract 16 (the black level) from the Y channel before processing, negative Y values should be supported after the subtraction. This allows the design to pass through real-world and test video signals with minimum artifacts.
Luma and Chroma Keying Keying involves specifying a desired foreground color; areas containing this color are replaced with a background image. Alternately, an area of any size or shape may be specified; foreground areas inside (or outside) this area are replaced with a background image.
Luminance Keying Luminance keying involves specifying a desired foreground luminance level; foreground areas containing luminance levels above (or below) the keying level are replaced with the background image. Alternately, this hard keying implementation may be replaced with soft keying by speci-
219
fying two luminance values of the foreground image: YH and YL (YL < YH). For keying the background into “white” foreground areas, foreground luminance values (YFG) above YH are replaced with the background image; YFG values below YL contain the foreground image. For YFG values between YL and YH, linear mixing is done between the foreground and background images. This operation may be expressed as: if YFG > YH K = 1 = background only if YFG < YL K = 0 = foreground only
if YH ≥ YFG ≥ YL K = (YFG – YL)/(YH – YL) = mix
By subtracting K from 1, the new luminance keying signal for keying into “black” foreground areas can be generated. Figure 7.17 illustrates luminance keying for two YCbCr sources. Although chroma keying typically uses a suppression technique to remove information from the foreground image, this is not done when luminance keying as the magnitudes of Cb and Cr are usually not related to the luminance level. Figure 7.18 illustrates luminance keying for R´G´B´ sources, which is more applicable for computer graphics. YFG may be obtained by the equation: YFG = 0.299R´ + 0.587G´ + 0.114B´
In some applications, the red and blue data is ignored, resulting in YFG being equal to only the green data. Figure 7.19 illustrates one technique of luminance keying between two digital composite video sources.
220
Chapter 7: Digital Video Processing
LUMINANCE KEY GENERATOR K 16
– BACKGROUND LUMINANCE (Y)
+
+
ROUNDING AND LIMITING
1–K
16
+
Y_OUT
16
– FOREGROUND LUMINANCE (Y)
+ MIXER K
128
– BACKGROUND CR
+
+
128
ROUNDING AND LIMITING
1–K
+
CR_OUT
128
– FOREGROUND CR
+ MIXER K
128
– BACKGROUND CB
+
+
128
1–K
ROUNDING AND LIMITING
+
128
– FOREGROUND CB
+ MIXER
Figure 7.17. Luminance Keying of Two YCbCr Video Signals.
CB_OUT
Luma and Chroma Keying
LUMINANCE KEY GENERATOR
K
BACKGROUND R
+
ROUNDING AND LIMITING
R_OUT
1–K
FOREGROUND R
MIXER K
BACKGROUND G
+
ROUNDING AND LIMITING
G_OUT
1–K
FOREGROUND G
MIXER K
BACKGROUND B
+
ROUNDING AND LIMITING
B_OUT
1–K
FOREGROUND B
MIXER
Figure 7.18. Luminance Keying of Two RGB Video Signals. RGB range is 0–255.
221
222
Chapter 7: Digital Video Processing
Y
LUMINANCE KEY GENERATOR
Y/C SEPARATOR
K
BLACK LEVEL
–
BACKGROUND VIDEO
+
+
BLACK LEVEL
1–K
ROUNDING AND LIMITING
+
OUT
BLACK LEVEL
–
FOREGROUND VIDEO
+ MIXER
Figure 7.19. Luminance Keying of Two Digital Composite Video Signals.
Chroma Keying Chroma keying involves specifying a desired foreground key color; foreground areas containing the key color are replaced with the background image. Cb and Cr are used to specify the key color; luminance information may be used to increase the realism of the chroma keying function. The actual mixing of the two video sources may be done in the component or composite domain, although component mixing reduces artifacts. Early chroma keying circuits simply performed a hard or soft switch between the foreground and background sources. In addition to limiting the amount of fine detail maintained in the foreground image, the background was not visible through transparent or translucent fore-
ground objects, and shadows from the foreground were not present in areas containing the background image. Linear keyers were developed that combine the foreground and background images in a proportion determined by the key level, resulting in the foreground image being attenuated in areas containing the background image. Although allowing foreground objects to appear transparent, there is a limit on the fineness of detail maintained in the foreground. Shadows from the foreground are not present in areas containing the background image unless additional processing is done— the luminance levels of specific areas of the background image must be reduced to create the effect of shadows cast by foreground objects.
Luma and Chroma Keying
If the blue or green backing used with the foreground scene is evenly lit except for shadows cast by the foreground objects, the effect on the background will be that of shadows cast by the foreground objects. This process, referred to as shadow chroma keying, or luminance modulation, enables the background luminance levels to be adjusted in proportion to the brightness of the blue or green backing in the foreground scene. This results in more realistic keying of transparent or translucent foreground objects by preserving the spectral highlights. Note that green backgrounds are now more commonly used due to lower chroma noise. Chroma keyers are also limited in their ability to handle foreground colors that are close to the key color without switching to the
223
background image. Another problem may be a bluish tint to the foreground objects as a result of blue light reflecting off the blue backing or being diffused in the camera lens. Chroma spill is difficult to remove since the spill color is not the original key color; some mixing occurs, changing the original key color slightly. One solution to many of the chroma keying problems is to process the foreground and background images individually before combining them, as shown in Figure 7.20. Rather than choosing between the foreground and background, each is processed individually and then combined. Figure 7.21 illustrates the major processing steps for both the foreground and background images during the chroma key process. Not shown in Figure 7.20 is the circuitry to initially subtract 16 (Y) or
FOREGROUND SUPPRESSOR
FOREGROUND VIDEO (YCBCR)
Y
KEY GENERATOR
KFG
FOREGROUND GAIN
NONADDITIVE MIX
KEY PROCESSOR
+
KBG GARBAGE MATTE
BACKGROUND GAIN
BACKGROUND VIDEO (YCBCR)
Figure 7.20. Typical Component Chroma Key Circuit.
OUTPUT VIDEO (YCBCR)
224
Chapter 7: Digital Video Processing
(A)
(B)
(C)
(D)
(E)
(F)
Figure 7.21. Major Processing Steps During Chroma Keying. (a) Original foreground scene. (b) Original background scene. (c) Suppressed foreground scene. (d) Background keying signal. (e) Background scene after multiplication by background key. (f) Composite scene generated by adding (c) and (e).
Luma and Chroma Keying
128 (Cb and Cr) from the foreground and background video signals and the addition of 16 (Y) or 128 (Cb and Cr) after the final output adder. Any DC offset not removed will be amplified or attenuated by the foreground and background gain factors, shifting the black level. The foreground key (KFG) and background key (KBG) signals have a range of 0 to 1. The garbage matte key signal (the term matte comes from the film industry) forces the mixer to output the foreground source in one of two ways. The first method is to reduce KBG in proportion to increasing KFG. This provides the advantage of minimizing black edges around the inserted foreground. The second method is to force the background to black for all nonzero values of the matte key, and insert the foreground into the background “hole.” This requires a cleanup function to remove noise around the black level, as this noise affects the background picture due to the straight addition process. The garbage matte is added to the foreground key signal (KFG) using a non-additive mixer (NAM). A nonadditive mixer takes the brighter of the two pictures, on a sample-bysample basis, to generate the key signal. Matting is ideal for any source that generates its own keying signal, such as character generators, and so on. The key generator monitors the foreground Cb and Cr data, generating the foreground keying signal, KFG. A desired key color is selected, as shown in Figure 7.22. The foreground Cb and Cr data are normalized (generating Cb´ and Cr´) and rotated θ degrees to generate the X and Z data, such that the positive X axis passes as close as possible to the desired key color. Typically, θ may be varied in 1° increments, and optimum chroma keying
225
occurs when the X axis passes through the key color. X and Z are derived from Cb and Cr using the equations: X = Cb´ cos θ + Cr´ sin θ
Z = Cr´ cos θ – Cb´ sin θ
Since Cb´ and Cr´ are normalized to have a range of ±1, X and Z have a range of ±1. The foreground keying signal (KFG) is generated from X and Z and has a range of 0–1: KFG = X – (|Z|/(tan (α/2))) KFG = 0 if X < (|Z|/(tan (α/2)))
where α is the acceptance angle, symmetrically centered about the positive X axis, as shown in Figure 7.23. Outside the acceptance angle, KFG is always set to zero. Inside the acceptance angle, the magnitude of KFG linearly increases the closer the foreground color approaches the key color and as its saturation increases. Colors inside the acceptance angle are further processed by the foreground suppressor. The foreground suppressor reduces foreground color information by implementing X = X – KFG, with the key color being clamped to the black level. To avoid processing Cb and Cr when KFG = 0, the foreground suppressor performs the operations: CbFG = Cb – KFG cos θ CrFG = Cr – KFG sin θ
where CbFG and CrFG are the foreground Cb and Cr values after key color suppression. Early implementations suppressed foreground information by multiplying Cb and Cr by a clipped version of the KFG signal. This, however, generated in-band alias components due
226
Chapter 7: Digital Video Processing
CR´ Z RED MAGENTA
YELLOW
CB´
θ BLUE KEY COLOR
X
GREEN CYAN
Figure 7.22. Rotating the Normalized Cb and Cr (Cb´ and Cr´) Axes by θ to Obtain the X and Z Axes, Such That the X Axis Passes Through the Desired Key Color (Blue in This Example).
Z
RED MAGENTA
KFG = 0
KFG = 0
YELLOW
KFG = 0.5 α/ 2
UNSUPPRESSED FOREGROUND COLORS
X = 0.5
α/ 2
BLUE SUPPRESSED FOREGROUND COLORS
X
KFG = 0.5
KFG = 0
GREEN CYAN
Figure 7.23. Foreground Key Values and Acceptance Angle.
Luma and Chroma Keying
to the multiplication and clipping process and produced a hard edge at key color boundaries. Unless additional processing is done, the CbFG and CrFG components are set to zero only if they are exactly on the X axis. Hue variations due to noise or lighting will result in areas of the foreground not being entirely suppressed. Therefore, a suppression angle is set, symmetrically centered about the positive X axis. The suppression angle (β) is typically configurable from a minimum of zero degrees, to a maximum of about one-third the acceptance angle (α). Any CbCr components that fall within this suppression angle are set to zero. Figure 7.24 illustrates the use of the suppression angle. Foreground luminance, after being normalized to have a range of 0–1, is suppressed by: YFG = Y´ – ySKFG YFG = 0 if ySKFG > Y´
Here, yS is a programmable value and used to adjust YFG so that it is clipped at the black level in the key color areas. The foreground suppressor also removes key-color fringes on wanted foreground areas caused by chroma spill, the overspill of the key color, by removing discolorations of the wanted foreground objects. Ultimatte® improves on this process by measuring the difference between the blue and green colors, as the blue backing is never pure blue and there may be high levels of blue in the foreground objects. Pure blue is rarely found in nature, and most natural blues have a higher content of green than red. For this reason, the red, green, and blue levels are monitored to differentiate between the blue backing and blue in wanted foreground objects. If the difference between blue and green is great enough, all three colors are set to zero to
227
produce black; this is what happens in areas of the foreground containing the blue backing. If the difference between blue and green is not large, the blue is set to the green level unless the green exceeds red. This technique allows the removal of the bluish tint caused by the blue backing while being able to reproduce natural blues in the foreground. As an example, a white foreground area normally would consist of equal levels of red, green, and blue. If the white area is affected by the key color (blue in this instance), it will have a bluish tint—the blue levels will be greater than the red or green levels. Since the green does not exceed the red, the blue level is made equal to the green, removing the bluish tint. There is a price to pay, however. Magenta in the foreground is changed to red. A green backing can be used, but in this case, yellow in the foreground is modified. Usually, the clamping is released gradually to increase the blue content of magenta areas. The key processor generates the initial background key signal (K´BG) used to remove areas of the background image where the foreground is to be visible. K´BG is adjusted to be zero in desired foreground areas and unity in background areas with no attenuation. It is generated from the foreground key signal (KFG) by applying lift (kL) and gain (kG) adjustments followed by clipping at zero and unity values: K´BG = (KFG – kL)kG
Figure 7.25 illustrates the operation of the background key signal generation. The transition between K´BG = 0 and K´BG = 1 should be made as wide as possible to minimize discontinuities in the transitions between foreground and background areas. For foreground areas containing the same CbCr values, but different luminance (Y) val-
228
Chapter 7: Digital Video Processing Z RED MAGENTA
YELLOW
KFG = 0
AFTER SUPPRESSION
BEFORE SUPPRESSION
α/ 2
BLUE
KFG = 0
X
COLOR SHIFTS AS A RESULT OF SUPPRESSION KFG = 0 GREEN CYAN
(A)
Z RED MAGENTA
KFG = 0 KFG = 0
YELLOW
COLOR SHIFTS AS A RESULT OF SUPPRESSION
BLUE
β/2
X
X = Z = 0 AFTER SUPPRESSION KFG = 0 GREEN CYAN
(B)
Figure 7.24. Suppression Angle Operation for a Gradual Change from a Red Foreground Object to the Blue Key Color. (a) Simple suppression. (b) Improved suppression using a suppression angle.
Luma and Chroma Keying
229
Z RED MAGENTA
K´BG = 0
K´BG = 0
YELLOW
K´BG = 0.5
KL K´BG = 1 1 / KG
BLUE X
K´BG = 1 K´BG = 0.5 K´BG = 0 GREEN CYAN
KEY COLOR
Figure 7.25. Background Key Generation.
ues, as the key color, the key processor may also reduce the background key value as the foreground luminance level increases, allowing turning off the background in foreground areas containing a “lighter” key color, such as light blue. This is done by: KBG = K´BG – ycYFG KBG = 0 if ycYFG > KFG
To handle shadows cast by foreground objects, and opaque or translucent foreground objects, the luminance level of the blue backing of the foreground image is monitored. Where the luminance of the blue backing is reduced, the luminance of the background image also is reduced. The amount of background luminance reduction must be controlled so that defects in the blue backing
(such as seams or footprints) are not interpreted as foreground shadows. Additional controls may be implemented to enable the foreground and background signals to be controlled independently. Examples are adjusting the contrast of the foreground so it matches the background or fading the foreground in various ways (such as fading to the background to make a foreground object vanish or fading to black to generate a silhouette). In the computer environment, there may be relatively slow, smooth edges—especially edges involving smooth shading. As smooth edges are easily distorted during the chroma keying process, a wide keying process is usually used in these circumstances. During wide keying, the keying signal starts before the edge of the graphic object.
230
Chapter 7: Digital Video Processing
Composite Chroma Keying In some instances, the component signals (such as YCbCr) are not directly available. For these situations, composite chroma keying may be implemented, as shown in Figure 7.26. To detect the chroma key color, the foreground video source must be decoded to produce the Cb and Cr color difference signals. The keying signal, KFG, is then used to mix between the two composite video sources. The garbage matte key signal forces the mixer to output the background source by reducing KFG.
Chroma keying using composite video signals usually results in unrealistic keying, since there is inadequate color bandwidth. As a result, there is a lack of fine detail, and halos may be present on edges. Superblack Keying Video editing systems also may make use of superblack keying. In this application, areas of the foreground composite video signal that have a level of 0 to –5 IRE are replaced with the background video information.
CB, CR
KEY GENERATOR
DECODE
GARBAGE MATTE
KFG
BACKGROUND VIDEO
+
OUTPUT VIDEO
–
FOREGROUND VIDEO
+
Figure 7.26. Typical Composite Chroma Key Circuit.
Video Scaling
Video Scaling With all the various video resolutions (Table 7.3), scaling is usually needed in almost every solution. When generating objects that will be displayed on SDTV, computer users must be concerned with such things as text size, line thickness, and so forth. For example, text readable on a 1280 × 1024 computer display may not be readable on a SDTV display due to the large amount of downscaling involved. Thin horizontal lines may either disappear completely or flicker at a 25- or 29.97-Hz rate when converted to interlaced SDTV.
Note that scaling must be performed on component video signals (such as R´G´B´ or YCbCr). Composite color video signals cannot be scaled directly due to the color subcarrier phase information present, which would be meaningless after scaling. In general, the spacing between output samples can be defined by a Target Increment (tarinc) value: tarinc = I / O
where I and O are the number of input (I) and output (O) samples, either horizontally or vertically. The first and last output samples may be aligned with the first and last input samples by adjusting the equation to be: tarinc = (I – 1) / (O – 1)
704 × 480
854 × 480 704 × 576
854 × 576
1280 × 720
1280 × 768
1920 × 1080
Displays
SDTV Sources
640 × 480
704 × 360
1024 × 768
528 × 480
800 × 600
1280 × 768 1366 × 768
1024 × 1024
1280 × 1024
231
1
480 × 480
544 × 480
640 × 480
704 × 480
704 × 432
1
480 × 576
544 × 576 704 × 576
768 × 576
HDTV Sources 1280 × 720
1440 × 8162
1440 × 10403 1280 × 1080
1440 × 1080
1920 × 1080
Table 7.3. Common Active Resolutions for Consumer Displays and Broadcast Sources. 116:9 letterbox on a 4:3 display. 22.35:1 anamorphic for a 16:9 1920x1080 display. 31.85:1 anamorphic for a 16:9 1920x1080 display.
232
Chapter 7: Digital Video Processing
Pixel Dropping and Duplication This is also called “nearest neighbor” scaling since only the input sample closest to the output sample is used. The simplest form of scaling down is pixel dropping, where (m) out of every (n) samples are thrown away both horizontally and vertically. A modified version of the Bresenham line-drawing algorithm (described in most computer graphics books) is typically used to determine which samples not to discard. Simple upscaling can be accomplished by pixel duplication, where (m) out of every (n) samples are duplicated both horizontally and vertically. Again, a modified version of the Bresenham line-drawing algorithm can be used to determine which samples to duplicate. Scaling using pixel dropping or duplication is not recommended due to the visual artifacts and the introduction of aliasing components.
Linear Interpolation An improvement in video quality of scaled images is possible using linear interpolation. When an output sample falls between two input samples (horizontally or vertically), the output sample is computed by linearly interpolating between the two input samples. However, scaling to images smaller than one-half of the original still results in deleted samples. Figure 7.27 illustrates the vertical scaling of a 16:9 image to fit on a 4:3 display. A simple bi-linear vertical filter is commonly used, as shown in Figure 7.28a. Two source samples, Ln and Ln+1, are weighted and added together to form a destination sample, Dm. D0 = 0.75L0 + 0.25L1 D1 = 0.5L1 + 0.5L2 D2 = 0.25L2 + 0.75L3
However, as seen in Figure 7.28a, this results in uneven line spacing, which may result in visual artifacts. Figure 7.28b illustrates vertical filtering that results in the output lines being more evenly spaced: D0 = L0 D1 = (2/3)L1 + (1/3)L2 D2 = (1/3)L2 + (2/3)L3
The linear interpolator is a poor bandwidth-limiting filter. Excess high-frequency detail is removed unnecessarily and too much energy above the Nyquist limit is still present, resulting in aliasing.
Anti-Aliased Resampling The most desirable approach is to ensure the frequency content scales proportionally with the image size, both horizontally and vertically. Figure 7.29 illustrates the fundamentals of an anti-aliased resampling process. The input data is upsampled by A and lowpass filtered to remove image frequencies created by the interpolation process. Filter B bandwidth-limits the signal to remove frequencies that will alias in the resampling process B. The ratio of B/A determines the scaling factor. Filters A and B are usually combined into a single filter. The response of the filter largely determines the quality of the interpolation. The ideal lowpass filter would have a very flat passband, a sharp cutoff at half of the lowest sampling frequency (either input or output), and very high attenuation in the stopband. However, since such a filter generates ringing on sharp edges, it is usually desirable to roll off the top of the passband. This makes for slightly softer pictures, but with less pronounced ringing.
Video Scaling
60
360 VISIBLE ACTIVE LINES
16:9 LETTERBOX PROGRAM
480 TOTAL ACTIVE LINES
(480) * (4 / 3) / (16 / 9) = 360
60
(A)
72
432 VISIBLE ACTIVE LINES
16:9 LETTERBOX PROGRAM
576 TOTAL ACTIVE LINES
(576) * (4 / 3) / (16 / 9) = 432
72
(B)
Figure 7.27. Vertical Scaling of 16:9 Images to Fit on a 4:3 Display. (a) 480-line systems. (b) 576-line systems.
233
234
Chapter 7: Digital Video Processing
L0
L0
0.75
D0
D0
0.25 L1
L1 2/3
0.5
D1 D1 1/3
0.5 L2
L2 1/3
0.25
L3
L4
D2
D2
0.75
2/3 L3
L4
0.75
D3
D3
0.25 L5
L5 2/3
0.5
D4
D4 1/3
0.5 L6
L6 1/3
0.25
L7
D5
0.75
D5
2/3 L7
(A)
(B)
Figure 7.28. 75% Vertical Scaling of 16:9 Images to Fit on a 4:3 Display. (a) Unevenly spaced results. (b) Evenly spaced results.
IN
UPSAMPLE BY A
LOWPASS FILTER A
LOWPASS FILTER B
RESAMPLE BY B
OUT
Figure 7.29. General Anti-Aliased Resampling Structure.
Scan Rate Conversion
Passband ripple and stopband attenuation of the filter provide some measure of scaling quality, but the subjective effect of ringing means a flat passband might not be as good as one might think. Lots of stopband attenuation is almost always a good thing. There are essentially three variations of the general resampling structure. Each combines the elements of Figure 7.29 in various ways. One approach is a variable-bandwidth antialiasing filter followed by a combined interpolator/resampler. In this case, the filter needs new coefficients for each scale factor—as the scale factor is changed, the quality of the image may vary. In addition, the overall response is poor if linear interpolation is used. However, the filter coefficients are time-invariant and there are no gain problems. A second approach is a combined filter/ interpolator followed by a resampler. Generally, the higher the order of interpolation, n, the better the overall response. The center of the filter transfer function is always aligned over the new output sample. With each scaling factor, the filter transfer function is stretched or compressed to remain aligned over n output samples. Thus, the filter coefficients, and the number of input samples used, change with each new output sample and scaling factor. Dynamic gain normalization is required to ensure the sum of the filter coefficients is always equal to one. A third approach is an interpolator followed by a combined filter/resampler. The input data is interpolated up to a common multiple of the input and output rates by the insertion of zero samples. This is filtered with a lowpass finite-impulse-response (FIR) filter to
235
interpolate samples in the zero-filled gaps, then re-sampled at the required locations. This type of design is usually achieved with a “polyphase” filter which switches its coefficients as the relative position of input and output samples change.
Display Scaling Examples Figures 7.30 through 7.38 illustrate various scaling examples for displaying 16:9 and 4:3 pictures on 4:3 and 16:9 displays, respectively. How content is displayed is a combination of user preferences and content aspect ratio. For example, when displaying 16:9 content on a 4:3 display, many users prefer to have the entire display filled with the cropped picture (Figure 7.31) rather than seeing black or gray bars with the letterbox solution (Figure 7.32). In addition, some displays incorrectly assume any progressive video signal on their YPbPr inputs is from an “anamorphic” source. As a result, they horizontally upscale progressive 16:9 programs by 25% when no scaling should be applied. Therefore, for set-top boxes it is useful to include a “16:9 (Compressed)” mode, which horizontally downscales the progressive 16:9 program by 25% to pre-compensate for the horizontally upscaling being done by the 16:9 display.
Scan Rate Conversion In many cases, some form of scan rate conversion (also called temporal rate conversion, frame rate conversion, or field rate conversion) is needed. Multi-standard analog VCRs and scan converters use scan rate conversion to
236
Chapter 7: Digital Video Processing
1920 Samples
1080 Scan Lines
Figure 7.30. 16:9 Source Example.
720 Samples
480 Scan Lines
Figure 7.31. Scaling 16:9 Content for a 4:3 Display: “Normal” or pan-and-scan mode. Results in some of the 16:9 content being ignored (indicated by gray regions).
Scan Rate Conversion
720 Samples
360 Scan Lines
480 Scan Lines
Figure 7.32. Scaling 16:9 Content for a 4:3 Display: “Letterbox” mode. Entire 16:9 program visible, with black bars at top and bottom of display.
720 Samples
480 Scan Lines
Figure 7.33. Scaling 16:9 Content for a 4:3 Display: “Squeezed” mode. Entire 16:9 program horizontally squeezed to fit 4:3 display, resulting in a distorted picture.
237
238
Chapter 7: Digital Video Processing
720 Samples
480 Scan Lines
Figure 7.34. 4:3 Source Example.
1920 Samples
1080 Scan Lines
Figure 7.35. Scaling 4:3 Content for a 16:9 Display: “Normal” mode. Left and right portions of 16:9 display not used, so made black or gray.
Scan Rate Conversion
1920 Samples
1080 Scan Lines
Figure 7.36. Scaling 4:3 Content for a 16:9 Display: “Wide” mode. Entire picture linearly scaled horizontally to fill 16:9 display, resulting in distorted picture unless used with anamorphic content.
1920 Samples
1080 Scan Lines
Figure 7.37. Scaling 4:3 Content for a 16:9 Display: “Zoom” mode. Top and bottom portion of 4:3 picture deleted, then scaled to fill 16:9 display.
239
240
Chapter 7: Digital Video Processing
1920 Samples
1080 Scan Lines
Figure 7.38. Scaling 4:3 Content for a 16:9 Display: “Panorama” mode. Left and right 25% edges of picture are nonlinearly scaled horizontally to fill 16:9 display, distorted picture on left and right sides.
convert between various video standards. Computers usually operate the display at about 75 Hz noninterlaced, yet need to display 50and 60-Hz interlaced video. With digital television, multiple refresh rates can be supported. Note that processing must be performed on component video signals (such as R´G´B´ or YCbCr). Composite color video signals cannot be processed directly due to the color subcarrier phase information present, which would be meaningless after processing.
Frame or Field Dropping and Duplicating Simple scan-rate conversion may be done by dropping or duplicating one out of every N fields. For example, the conversion of 60-Hz to 50-Hz interlaced operation may drop one out of
every six fields, as shown in Figure 7.39, using a single field store. The disadvantage of this technique is that the viewer may be see jerky motion, or motion “judder.” In addition, some MPEG decoders use top-field only to convert from 60Hz to 50Hz, degrading the vertical resolution. The worst artifacts are present when a non-integer scan rate conversion is done—for example, when some frames are displayed three times, while others are displayed twice. In this instance, the viewer will observe double or blurred objects. As the human brain tracks an object in successive frames, it expects to see a regular sequence of positions, and has trouble reconciling the apparent stop-start motion of objects. As a result, it incorrectly concludes that there are two objects moving in parallel.
Scan Rate Conversion
FIELD 1
2
3
4
5
6
60 HZ INTERLACED (WRITE)
FIELD 1
2
3
4
5
50 HZ INTERLACED (READ)
Figure 7.39. 60-Hz to 50-Hz Conversion Using a Single Field Store by Dropping One out of Every Six Fields.
1
2
3
4
5
6
525 / 50 FIELDS 17 %
100 %
33 % 83 %
50 % 67 %
67 % 50 %
83 % 33 %
100 %
17 %
525 / 60 FIELDS 1
2
3
4
5
6
7
Figure 7.40. 50-Hz to 60-Hz Conversion Using Temporal Interpolation with No Motion Compensation.
241
242
Chapter 7: Digital Video Processing
Temporal Interpolation This technique generates new frames from the original frames as needed to generate the desired frame rate. Information from both past and future input frames should be used to optimally handle objects appearing and disappearing. Conversion of 50-Hz to 60-Hz operation using temporal interpolation is illustrated in Figure 7.40. For every five fields of 50-Hz video, there are six fields of 60-Hz video. After both sources are aligned, two adjacent 50-Hz fields are mixed together to generate a new 60-Hz field. This technique is used in some inexpensive standards converters to convert between 625/50 and 525/60 standards. Note that no motion analysis is done. Therefore, if the camera operating at 625/50 pans horizontally past a narrow vertical object, you see one object once every six 525/60 fields, and for the five fields in between, you see two objects, one fading in while the other fades out. 625/50 to 525/60 Examples Figure 7.41 illustrates a scan rate converter that implements vertical, followed by temporal, interpolation. Figure 7.42 illustrates the spectral representation of the design in Figure 7.41. Many designs now combine the vertical and temporal interpolation into a single design, as shown in Figure 7.43, with the corresponding spectral representation shown in Figure 7.44. This example uses vertical, followed by temporal, interpolation. If temporal, followed by vertical, interpolation were implemented, the field stores would be half the size. However, the number of line stores would increase from four to eight. In either case, the first interpolation process must produce an intermediate, higher-
resolution progressive format to avoid interlace components that would interfere with the second interpolation process. It is insufficient to interpolate, either vertically or temporally, using a mixture of lines from both fields, due to the interpolation process not being able to compensate for the temporal offset of interlaced lines. Motion Compensation Higher-quality scan rate converters using temporal interpolation incorporate motion compensation to minimize motion artifacts. This results in extremely smooth and natural motion, and images appear sharper and do not suffer from motion “judder.” Motion estimation for scan rate conversion differs from that used by MPEG. In MPEG, the goal is to minimize the displaced frame difference (error) by searching for a high correlation between areas in subsequent frames. The resulting motion vectors do not necessarily correspond to true motion vectors. For scan rate conversion, it is important to determine true motion information to perform correct temporal interpolation. The interpolation should be tolerant of incorrect motion vectors to avoid introducing artifacts as unpleasant as those the technique is attempting to remove. Motion vectors could be incorrect for several reasons, such as insufficient time to track the motion, out-of-range motion vectors, and estimation difficulties due to aliasing. 100 Hz Interlaced Television Example A standard PAL television shows 50 fields per second. The images flicker, especially when you look at large areas of highly-saturated color. A much improved picture can be achieved using a 100 Hz interlaced refresh (also called double scan).
Scan Rate Conversion
243
VERTICAL INTERPOLATOR
625 / 50 INTERLACED
H
H
H
H
+
+
+
+
+
+ 525 / 50 SEQUENTIAL
F
F
F
F
+
+
+
525 / 60 INTERLACED
TEMPORAL INTERPOLATOR
F = FIELD STORE H = LINE STORE
Figure 7.41. Typical 625/50 to 525/60 Conversion Using Vertical, Followed by Temporal, Interpolation.
244
Chapter 7: Digital Video Processing
VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
937.5 787.5
625 525
312.5 262.5
0
0 0
25
50
0
75
25
50
75
TEMPORAL FREQUENCY (HZ)
TEMPORAL FREQUENCY (HZ)
(B)
(A) VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
787.5
525
262.5
0
0
30
60
TEMPORAL FREQUENCY (HZ)
(C)
Figure 7.42. Spectral Representation of Vertical, Followed by Temporal, Interpolation. (a) Vertical lowpass filtering. (b) Resampling to intermediate sequential format and temporal lowpass filtering. (c) Resampling to final standard.
+ + + +
F = FIELD STORE H = LINE STORE A
+
H 625 / 50 INTERLACED
H
+
H H
H
F F
A
H
F
A
H
F
A
525 / 60 INTERLACED
Scan Rate Conversion
Figure 7.43. Typical 625/50 to 525/60 Conversion Using Combined Vertical and Temporal Interpolation.
245
246
Chapter 7: Digital Video Processing
VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
937.5 787.5
625 525
312.5 262.5
0
0
0
25
50
75
TEMPORAL FREQUENCY (HZ)
(A)
0
30
60
TEMPORAL FREQUENCY (HZ)
(B)
Figure 7.44. Spectral Representation of Combined Vertical and Temporal Interpolation. (a) Two-dimensional lowpass filtering. (b) Resampling to final standard.
Scan Rate Conversion
50 HZ SOURCE TIME
REPEATED FIELDS
100 HZ DISPLAY (A)
100 HZ DISPLAY (B)
CALCULATED FIELDS
Figure 7.45. 50 Hz to 100 Hz (Double Scan Interlaced) Techniques.
247
248
Chapter 7: Digital Video Processing
Early 100 Hz televisions simply repeated fields (F1F1F2F2F3F3F4F4...), as shown in Figure 7.45a. However, they still had line flicker, where horizontal lines constantly jumped between the odd and even lines. This disturbance occurred once every twenty-fifth of a second. The field sequence F1F2F1F2F3F4F3F4... can be used, which solves the line flicker problem. Unfortunately, this gives rise to the problem of judder in moving images. This can be compensated for by using the F1F2F1F2F3F4F3F4... sequence for static images, and the F1F1F2F2F3F3F4F4... sequence for moving images. An ideal picture is still not obtained when viewing programs created for film. They are subject to judder, owing to the fact that each film frame is transmitted twice. Instead of the field sequence F1F1F2F2F3F3F4F4..., the situation calls for the sequence F1F1´F2F2´F3F3´F4F4´... (Figure 7.45b), where Fn´ is a motion-compensated generated image between Fn and Fn+1.
pled during decoding to restore the original pitch (costly to do in a low-cost consumer product) or resampling may be done during the program authoring. One of these two solutions must be used since many audio decoders cannot handle the 4% faster audio data via S/PDIF (IEC 60958).
3:2 Pulldown When converting 24 frames per second content to NTSC (59.94-Hz field rate), 3:2 pulldown is commonly used, as shown in Figure 7.46. The film speed is slowed down by 0.1% to 23.976 (24/1.001) frames per second. Two film frames generate five video fields. In scenes of high-speed motion of objects, the specific film frame used for a particular video field may be manually adjusted to minimize motion artifacts. FILM FRAME O
N
E
VIDEO FIELD
WHITE FLAG
N
W
N + 1
O
N + 2
2:2 Pulldown This technique is used with some filmbased MPEG content for 50 Hz regions. Film is usually recorded at 24 frames per second. During MPEG encoding, the telecine machine is sped up from 24 to 25 frames per second, making the content 25 frames per second progressive. During MPEG decoding, each film frame is simply mapped into two video fields. This technique provides higher video quality and avoids motion judder artifacts. However, it shortens the duration of the program by about 4 percent, cutting the duration of a two-hour movie by ~5 minutes. To compensate the audio changing pitch due to the telecine speedup, it may be resam-
E
N + 1
N + 3
W
O
N + 4 E
N + 2
O
N + 5
W
N + 6
E N + 7 O
N + 3
N + 8
W
E N + 9
O = ODD LINES OF FILM FRAME E = EVEN LINES OF FILM FRAME
Figure 7.46. Typical 3:2 Pulldown for Transferring Film to NTSC Video.
Noninterlaced-to-Interlaced Conversion
3:2 pulldown may also be used by MPEG decoders simply to increase the frame rate from 23.976 (24/1.001) to 59.94 (60/1.001) frames per second, avoiding the deinterlacing issue. Varispeed is commonly used to cover up problems such as defects, splicing, censorship cuts, or to change the running time of a program. Rather than repeating film frames and causing a “stutter,” the 3:2 relationship between the film and video is disrupted long enough to ensure a smooth temporal rate. Analog laserdiscs use a white flag signal to indicate the start of another sequence of related fields for optimum still-frame performance. During still-frame mode, the white flag signal tells the system to back up two fields (to use two fields that have no motion between them) to re-display the current frame.
3:3 Pulldown This technique is used in some displays that support 72 Hz refresh. The 24 frames per second film-based content is converted to 72 Hz progressive by simply duplicating each film frame three times.
24:1 Pulldown This technique, also called “12:1 pulldown,” can also be used to convert 24 frames/ second content to 50 fields per second. Two video fields are generated from every film frame, except every 12th film frame generates 3 video fields. Although the audio pitch is correct, motion judder is present every onehalf second when smooth motion is present.
249
Noninterlaced-to-Interlaced Conversion In some applications, it is necessary to display a noninterlaced video signal on an interlaced display. Thus, some form of “noninterlaced-to-interlaced conversion” may be required. Noninterlaced to interlaced conversion must be performed on component video signals (such as R´G´B´ or YCbCr). Composite color video signals (such as NTSC or PAL) cannot be processed directly due to the presence of color subcarrier phase information, which would be meaningless after processing. These signals must be decoded into component color signals, such as R´G´B´ or YCbCr, prior to conversion. There are essentially two techniques: scan line decimation and vertical filtering.
Scan Line Decimation The easiest approach is to throw away every other active scan line in each noninterlaced frame, as shown in Figure 7.47. Although the cost is minimal, there are problems with this approach, especially with the top and bottom of objects. If there is a sharp vertical transition of color or intensity, it will flicker at one-half the refresh rate. The reason is that it is only displayed every other field as a result of the decimation. For example, a horizontal line that is one noninterlaced scan line wide will flicker on and off. Horizontal lines that are two noninterlaced scan lines wide will oscillate up and down. Simple decimation may also add aliasing artifacts. While not necessarily visible, they will affect any future processing of the picture.
250
Chapter 7: Digital Video Processing
NONINTERLACED FRAME N
1
INTERLACED FIELD 1
1
2 3
6 7 8
NONINTERLACED ACTIVE LINE NUMBER
INTERLACED ACTIVE LINE NUMBER
2
5 6
4
1
3 4
3
INTERLACED FIELD 2
1 2
2
4 5
NONINTERLACED FRAME N + 1
3
7 8
4
NONINTERLACED ACTIVE LINE NUMBER
INTERLACED ACTIVE LINE NUMBER
Figure 7.47. Noninterlaced-to-Interlaced Conversion Using Scan Line Decimation.
NONINTERLACED FRAME N
1
INTERLACED FIELD 1
1
2 3
6 7 8
NONINTERLACED ACTIVE LINE NUMBER
INTERLACED ACTIVE LINE NUMBER
2
5 6
4
1
3 4
3
INTERLACED FIELD 2
1 2
2
4 5
NONINTERLACED FRAME N + 1
3
7 8
4
NONINTERLACED ACTIVE LINE NUMBER
INTERLACED ACTIVE LINE NUMBER
Figure 7.48. Noninterlaced-to-Interlaced Conversion Using 3-Line Vertical Filtering.
Interlaced-to-Noninterlaced Conversion
Vertical Filtering A better solution is to use two or more lines of noninterlaced data to generate one line of interlaced data. Fast vertical transitions are smoothed out over several interlaced lines. For a 3-line filter, such as shown in Figure 7.48, typical coefficients are [0.25, 0.5, 0.25]. Using more than three lines usually results in excessive blurring, making small text difficult to read. An alternate implementation uses IIR rather than FIR filtering. In addition to averaging, this technique produces a reduction in brightness around objects, further reducing flicker. Note that care must be taken at the beginning and end of each frame in the event that fewer scan lines are available for filtering.
Interlaced-to-Noninterlaced Conversion In some applications, it is necessary to display an interlaced video signal on a noninterlaced display. Thus, some form of “deinterlacing” or “progressive scan conversion” may be required. Note that deinterlacing must be performed on component video signals (such as R´G´B´ or YCbCr). Composite color video signals (such as NTSC or PAL) cannot be deinterlaced directly due to the presence of color subcarrier phase information, which would be meaningless after processing. These signals must be decoded into component color signals, such as R´G´B´ or YCbCr, prior to deinterlacing. There are two fundamental deinterlacing algorithms: video mode and film mode. Video mode deinterlacing can be further broken down into interfield and intrafield processing.
251
The goal of a good deinterlacer is to correctly choose the best algorithm needed at a particular moment. In systems where the vertical resolution of the source and display do not match (due to, for example, displaying SDTV content on a HDTV), the deinterlacing and vertical scaling can be merged into a single process.
Video Mode: Intrafield Processing This is the simplest method for generating additional scan lines using only information in the original field. The computer industry has coined this technique as “bob.” Although there are two common techniques for implementing intrafield processing, scan line duplication and scan line interpolation, the resulting vertical resolution is always limited by the content of the original field. Scan Line Duplication Scan line duplication (Figure 7.49) simply duplicates the previous active scan line. Although the number of active scan lines is doubled, there is no increase in the vertical resolution. Scan Line Interpolation Scan line interpolation generates interpolated scan lines between the original active scan lines. Although the number of active scan lines is doubled, the vertical resolution is not. The simplest implementation, shown in Figure 7.50, uses linear interpolation to generate a new scan line between two input scan lines: outn = (inn–1 + inn+1) / 2
Better results, at additional cost, may be achieved by using a FIR filter:
252
Chapter 7: Digital Video Processing
INPUT FIELD ACTIVE LINES
OUTPUT FRAME ACTIVE LINES
INPUT FIELD ACTIVE LINES
1
1
OUTPUT FRAME ACTIVE LINES
1
1
2 = 1
2 = (1 + 3) / 2
3
2
3
2
4 = 3
4 = (3 + 5) / 2
5
3
5
3
6 = 5
6 = (5 + 7) / 2
7
4
7
4
8 = 7
8 = (7 + 9) / 2
Figure 7.49. Deinterlacing Using Scan Line Duplication. New scan lines are generated by duplicating the active scan line above it.
FIELD 1 ACTIVE LINE
FIELD 2 ACTIVE LINE
1
Figure 7.50. Deinterlacing Using Scan Line Interpolation. New scan lines are generated by averaging the previous and next active scan lines.
DEINTERLACED FRAME ACTIVE LINE
1 1
2 3
2 2
2
4
1
5
3 3
6
4
8
4
7
Figure 7.51. Deinterlacing Using Field Merging. Shaded scan lines are generated by using the input scan line from the next or previous field.
2 1
INPUT FIELD NUMBER
4 3
4 3
6 5
6 5
8 7
8 7
9
10 9
OUTPUT FRAME NUMBER
Figure 7.52. Producing Deinterlaced Frames at Field Rates.
Interlaced-to-Noninterlaced Conversion
outn = (160*(inn–1 + inn+1) – 48*(inn–3 + inn+3) + 24*(inn–5 + inn+5) – 12*(inn–7 + inn+7) + 6*(inn–9 + inn+9) – 2*(inn–11 + inn+11)
Fractional Ratio Interpolation In many cases, there is a periodic, but nonintegral, relationship between the number of input scan lines and the number of output scan lines. In this case, fractional ratio interpolation may be necessary, similar to the polyphase filtering used for scaling only performed in the vertical direction. This technique combines deinterlacing and vertical scaling into a single process. Variable Interpolation In a few cases, there is no periodicity in the relationship between the number of input and output scan lines. Therefore, in theory, an infiOBJECT POSITION IN FIELD ONE
253
nite number of filter phases and coefficients are required. Since this is not feasible, the solution is to use a large, but finite, number of filter phases. The number of filter phases determines the interpolation accuracy. This technique also combines deinterlacing and vertical scaling into a single process.
Video Mode: Interfield Processing In this method, video information from more than one field is used to generate a single progressive frame. This method can provide higher vertical resolution since it uses content from more than a single field. Field Merging This technique merges two consecutive fields together to produce a frame of video (Figure 7.51). At each field time, the active scan lines of that field are merged with the active scan lines of the previous field. The OBJECT POSITION IN FIELD TWO
OBJECT POSITIONS IN MERGED FIELDS
Figure 7.53. Movement Artifacts When Field Merging Is Used.
254
Chapter 7: Digital Video Processing
result is that for each input field time, a pair of fields combine to generate a frame (see Figure 7.52). Although simple to implement, the vertical resolution is doubled only in regions of no movement. Moving objects will have artifacts, also called “combing,” due to the time difference between two fields—a moving object is located in a different position from one field to the next. When the two fields are merged, moving objects will have a “double image” (see Figure 7.53). It is common to soften the image slightly in the vertical direction to attempt to reduce the visibility of combing. When implemented, it causes a loss of vertical resolution and jitter on movement and pans. The computer industry refers to this technique as “weave,” but “weave” also includes the inverse telecine process to remove any 3:2 pulldown present in the source. Theoretically, this eliminates the “double image” artifacts since two identical fields are now being merged. Motion Adaptive Deinterlacing A good deinterlacing solution is to use field merging for still areas of the picture and scan line interpolation for areas of movement. To accomplish this, motion, on a sample-by-sample basis, must be detected over the entire picture in real time, requiring processing several fields of video. As two fields are combined, full vertical resolution is maintained in still areas of the picture, where the eye is most sensitive to detail. The sample differences may have any value, from 0 (no movement and noise-free) to maximum (for example, a change from full intensity to black). A choice must be made when to use a sample from the previous field (which is in the wrong location due to motion) or to interpolate a new sample from adjacent scan lines in
the current field. Sudden switching between methods is visible, so crossfading (also called soft switching) is used. At some magnitude of sample difference, the loss of resolution due to a double image is equal to the loss of resolution due to interpolation. That amount of motion should result in the crossfader being at the 50% point. Less motion will result in a fade towards field merging and more motion in a fade towards the interpolated values. Rather than “per pixel” motion adaptive deinterlacing, which makes decisions for every sample, some low-cost solutions use “per field” motion adaptive deinterlacing. In this case, the algorithm is selected each field, based on the amount of motion between the fields. “Per pixel” motion adaptive deinterlacing, although difficult to implement, looks quite good when properly done. “Per field” motion adaptive deinterlacing rarely looks much better than vertical interpolation. Motion Compensated Deinterlacing Motion compensated (or motion vector steered) deinterlacing is several orders of magnitude more complex than motion adaptive deinterlacing, and is commonly found in provideo format converters. Motion compensated processing requires calculating motion vectors between fields for each sample, and interpolating along each sample’s motion trajectory. Motion vectors must also be found that pass through each of any missing samples. Areas of the picture may be covered or uncovered as you move between frames. The motion vectors must also have sub-pixel accuracy, and be determined in two temporal directions between frames. The motion vector errors used by MPEG are self-correcting since the residual difference between the predicted macroblocks is encoded. As motion compensated deinterlacing is a single-ended system, motion vector
Interlaced-to-Noninterlaced Conversion
errors will produce artifacts, so different search and verification algorithms must be used.
Film Mode (using Inverse Telecine) For sources that have 3:2 pulldown (i.e., 60 fields/second video converted from 24 frames/second film), higher deinterlacing performance may be obtained by removing duplicate fields prior to processing. The inverse telecine process detects the 3:2 field sequence and the redundant third fields are removed. The remaining field pairs are merged (since there is no motion between them) to form progressive frames at 24 frames/second. These are then repeated in a 3:2 sequence to get to 60 frames/second. Although this may seem to be the ideal solution, many MPEG-based sources use both 60 fields/second video (that has 3:2 pulldown) and 24 frames/second video (film-based) within a program. In addition, some programs may occasionally have both video types present simultaneously. In other cases, the 3:2 pulldown timing (cadence) doesn’t stay regular, or the source was never originally from film. Thus, the deinterlacer has to detect each video type and process it differently (video mode vs. film mode). Display artifacts are common due to the delay between the video type changing and the deinterlacer detecting the change.
Frequency Response Considerations Various two-times vertical upsampling techniques for deinterlacing may be implemented by stuffing zero values between two valid lines and filtering, as shown in Figure 7.54.
255
Line A shows the frequency response for line duplication, in which the lowpass filter coefficients for the filter shown are 1, 1, and 0. Line interpolation, using lowpass filter coefficients of 0.5, 1.0, and 0.5, results in the frequency response curve of Line B. Note that line duplication results in a better high-frequency response. Vertical filters with a better frequency response than the one for line duplication are possible, at the cost of more line stores and processing.
GAIN
C
1
A B
525
VERTICAL FREQUENCY (CYCLES PER PICTURE HEIGHT)
H
H
+
Figure 7.54. Frequency Response of Various Deinterlacing Filters. (a) Line duplication. (b) Line interpolation. (c) Field merging.
256
Chapter 7: Digital Video Processing
The best vertical frequency response is obtained when field merging is implemented. The spatial position of the lines is already correct and no vertical processing is required, resulting in a flat curve (Line C). Again, this applies only for stationary areas of the image.
DCT-Based Compression The transform process of many video compression standards is based on the Discrete Cosine Transform, or DCT. The easiest way to envision it is as a filter bank with all the filters computed in parallel. During encoding, the DCT is usually followed by several other operations, such as quantization, zig-zag scanning, run-length encoding, and variable-length encoding. During decoding, this process flow is reversed. Many times, the terms macroblocks and blocks are used when discussing video compression. Figure 7.55 illustrates the relationship between these two terms, and shows why transform processing is usually done on 8 × 8 samples.
DCT
The 8 × 8 DCT processes an 8 × 8 block of samples to generate an 8 × 8 block of DCT coefficients, as shown in Figure 7.56. The input may be samples from an actual frame of video or motion-compensated difference (error) values, depending on the encoder mode of operation. Each DCT coefficient indicates the amount of a particular horizontal or vertical frequency within the block. DCT coefficient (0,0) is the DC coefficient, or average sample value. Since natural images tend to vary only slightly from sample to sample, low frequency coefficients are typically
larger values and high frequency coefficients are typically smaller values. The 8 × 8 DCT is defined in Figure 7.57. f(x, y) denotes sample (x, y) of the 8 × 8 input block and F(u,v) denotes coefficient (u, v) of the DCT transformed block. A reconstructed 8 × 8 block of samples is generated using an 8 × 8 inverse DCT (IDCT), defined in Figure 7.58. Although exact reconstruction is theoretically achievable, it is not practical due to finite-precision arithmetic, quantization and differing IDCT implementations. As a result, there are “mismatches” between different IDCT implementations. “Mismatch control” attempts to reduce the drift between encoder and decoder IDCT results by eliminating bit patterns having the greatest contribution towards mismatches. MPEG-1 mismatch control is known as “oddification” since it forces all quantized DCT coefficients to negative values. MPEG-2 and MPEG-4 Part 2 use an improved method called “LSB toggling” which affects only the LSB of the 63rd DCT coefficient after inverse quanitzation. H.264 (also known as MPEG-4 Part 10) neatly sidesteps the issue by using an “exactmatch inverse transform.” Every decoder will produce exactly the same pictures, all else being equal.
Quantization
The 8 × 8 block of DCT coefficients is quantized, which reduces the overall precision of the integer coefficients and tends to eliminate high frequency coefficients, while maintaining perceptual quality. Higher frequencies are usually quantized more coarsely (fewer values allowed) than lower frequencies, due to visual perception of quantization error. The quantizer is also used for constant bit rate
DCT-Based Compression
DIVIDE PICTURE INTO 16 X 16 BLOCKS (MACROBLOCKS)
EACH MACROBLOCK IS 16 SAMPLES BY 16 LINES (4 BLOCKS) EACH BLOCK IS 8 SAMPLES BY 8 LINES
Figure 7.55. The Relationship between Macroblocks and Blocks.
FREQUENCY COEFFICIENTS DC TERM INCREASING HORIZONTAL FREQUENCY
DCT ISOLATED HIGH–FREQUENCY TERM
8 X 8 BLOCK
INCREASING VERTICAL FREQUENCY
Figure 7.56. The DCT Processes the 8 × 8 Block of Samples or Error Terms to Generate an 8 × 8 Block of DCT Coefficients.
257
258
Chapter 7: Digital Video Processing
F ( u, v ) = 0.25C ( u )C ( v ) ∑ ∑ f ( x, y ) cos ( ( ( 2x + 1 )uπ ) ⁄ 16 ) cos ( ( ( 2y + 1 )vπ ) ⁄ 16 ) 7
7
x = 0y = 0
u, v, x, y = 0, 1, 2, . . . 7 (x, y) are spatial coordinates in the sample domain (u, v) are coordinates in the transform domain
Figure 7.57. 8 × 8 Two-Dimensional DCT Definition. f ( x, y ) = 0.25 ∑ ∑ C ( u )C ( v )F ( u, v ) cos ( ( ( 2x + 1 )uπ ) ⁄ 16 ) cos ( ( ( 2y + 1 )vπ ) ⁄ 16 ) 7
7
u = 0v = 0
Figure 7.58. 8 × 8 Two-Dimensional Inverse DCT (IDCT) Definition.
applications where it is varied to control the output bit rate.
Zig-Zag Scanning The quantized DCT coefficients are rearranged into a linear stream by scanning them in a zig-zag order. This rearrangement places the DC coefficient first, followed by frequency coefficients arranged in order of increasing frequency, as shown in Figures 7.59, 7.60, and 7.61. This produces long runs of zero coefficients.
Run Length Coding The linear stream of quantized frequency coefficients is converted into a series of [run, amplitude] pairs. [run] indicates the number of zero coefficients, and [amplitude] the nonzero coefficient that ended the run.
Variable-Length Coding The [run, amplitude] pairs are coded using a variable-length code, resulting in additional lossless compression. This produces shorter codes for common pairs and longer codes for less common pairs. This coding method produces a more compact representation of the DCT coefficients, as a large number of DCT coefficients are usually quantized to zero and the re-ordering results (ideally) in the grouping of long runs of consecutive zero values.
DCT-Based Compression
0
1
5
6
14
15
27
28
2
4
7
13
16
26
29
42
3
8
12
17
25
30
41
43
9
11
18
24
31
40
44
53
10
19
23
32
39
45
52
54
20
22
33
38
46
51
55
60
21
34
37
47
50
56
59
61
35
36
48
49
57
58
62
63
A
F
LINEAR ARRAY OF 64 FREQUENCY COEFFICIENTS
ZIG–ZAG SCAN OF 8 X 8 BLOCK OF QUANTIZED FREQUENCY COEFFICIENTS
Figure 7.59. The 8 × 8 Block of Quantized DCT Coefficients Are Zig-Zag Scanned to Arrange in Order of Increasing Frequency. This scanning order is used for H.261, H.263, MPEG-1, MPEG-2, MPEG-4 Part 2, ITU-R BT.1618, ITU-R BT.1620, SMPTE 314M and SMPTE 370M.
0
4
6
20
22
36
38
52
1
5
7
21
23
37
39
53
2
8
19
24
34
40
50
54
3
9
18
25
35
41
51
55
10
17
26
30
42
46
56
60
11
16
27
31
43
47
57
61
12
15
28
32
44
48
58
62
13
14
29
33
45
49
59
63
A
F
LINEAR ARRAY OF 64 FREQUENCY COEFFICIENTS
ZIG–ZAG SCAN OF 8 X 8 BLOCK OF QUANTIZED FREQUENCY COEFFICIENTS
Figure 7.60. H.263, MPEG-2 and MPEG-4 Part 2 “Alternate-Vertical” Scanning Order.
259
260
Chapter 7: Digital Video Processing
0
1
2
3
10
11
12
13
4
5
8
9
17
16
15
14
6
7
19
18
26
27
28
29
20
21
24
25
30
31
32
33
22
23
34
35
42
43
44
45
36
37
40
41
46
47
48
49
38
39
50
51
56
57
58
59
52
53
54
55
60
61
62
63
A
F
LINEAR ARRAY OF 64 FREQUENCY COEFFICIENTS
ZIG–ZAG SCAN OF 8 X 8 BLOCK OF QUANTIZED FREQUENCY COEFFICIENTS
Figure 7.61. H.263 and MPEG-4 Part 2 “Alternate-Horizontal” Scanning Order.
Fixed Pixel Display Considerations
Detail Correction
The unique designs and color reproduction gamuts of fixed pixel displays have resulted in new video processing technologies being developed. The result is brighter, sharper, more colorful images regardless of the video source.
In CRT-based televisions, enhancing the image is commonly done by altering the electron beam diameter. With fixed-pixel displays, adding overshoot and undershoot to the video signals cause distortion. An acceptable implementation is to gradually change the brightness of the images before and after regions needing contour enhancement.
Expanded Color Reproduction
Non-uniform Quantization
Broadcast stations are usually tuned to meet the limited color reproduction characteristics of CRT-based televisions. To fit the color reproduction capabilities of PDP and LCD, manufacturers have introduced various color expansion technologies. These include using independent hue and saturation controls for each primary and complementary color, plus the flesh color.
Rather than simply increasing the number of quantization levels, the quantization steps can be changed in accordance with the intensity of the image. This is possible since people better detect small changes in brightness for dark images than for bright images. In addition, the brighter the image, the less sensitive people are to changes in brightness. This means that more quantization steps can be
Application Example
used for dark images than for bright ones. This technique can also be used to increase the quantization steps for shades that appear frequently.
Scaling and Deinterlacing Fixed-pixel displays, such as LCD and plasma, usually upscale then downscale during deinterlacing to minimize moiré noise due to folded distortion. For example, a 1080i source is deinterlaced to 2160p, scaled to 1536p, then finally scaled to 768p (to drive a 1024x768 display). Alternately, some solutions deinterlace and upscale to 1500p, then scale to the display's native resolution.
Application Example Figures 7.62 and 7.63 illustrate the typical video processing done after MPEG decoding and deinterlacing. In addition to the primary video source (such as a MPEG decoder), additional video sources typically include an on-screen-display (OSD), second video for picture-in-picture (PIP), graphics and closed captioning or subtitles. OSD design is unique to each product, so the OSD memory usually supports a wide variety of RGB/YCbCr formats and resolutions. Optional lookup tables add gamma correction to RGB data, convert 2-, 4- or 8-indexed color to 32-bit YCbCrA data, or translate 0–255 video levels to 16–235.
261
Graphics memory is usually used for JPEG images, background images, etc. It also usually supports a wide variety of RGB/YCbCr formats and resolutions. Optional lookup tables add gamma correction to RGB data, convert 2-, 4- or 8-indexed color to 32-bit YCbCrA data, or translate 0–255 video levels to 16–235. The subtitle memory is a useful region for rendering closed captioning, DVB subtitles and DVD subpictures. Lookup tables convert 2-, 4- or 8-indexed color to 32-bit YCbCrA data. Being able to scale each source independently offers maximum flexibility in the design of the display. In addition to being able to output any resolution regardless of the source resolutions, special effects can also be accommodated. Chromaticity correction circuits ensure colors are accurate independent of the sources and display (SDTV vs. HDTV). Independent brightness, contrast, saturation, hue and sharpness controls for each source and video output interface offer users the most flexibility. PIP can be adjusted without affecting the main picture, video can be adjusted without affecting still picture video quality, etc. The optional downscaling and progressiveto-interlaced conversion block for the top NTSC/PAL encoder in Figure 7.63 enables simultaneous HD and SD, or simultaneous progressive and interlaced, outputs without affecting the HD or progressive video quality. The second NTSC/PAL encoder shown at the bottom of Figure 7.63 is useful for recording a program without any OSD or subtitle information being accidently recorded.
262 Chapter 7: Digital Video Processing
24-BIT 4:4:4 YCBCR PLUS 8-BIT ALPHA
Figure 7.62. Video Composition Simplified Block Diagram.
OSD ON-SCREEN
PLANE
MENUS PRIORITY 1
SUBPICTURES, SUBTITLES, CAPTIONING
PIP VIDEO, JPEG IMAGES, GRAPHICS
RGB GAMMA
X -Y
4 x 256 x 8 LUT
SCALING
RGB --> YCBCR
WITH
NDX --> YCBCR
FLICKER
256 x 32 LUT
FILTER
COLORIMETRY CORRECT
SUBTITLE PLANE
NDX --> YCBCR
X -Y
256 x 32 LUT
SCALING
PRIORITY 2
PLANE
4 x 256 x 8 LUT
DECRYPTION
RGB --> YCBCR NDX --> YCBCR
PRIORITY 3
TO OUTPUT
MIX
FORMATTING
BRIGHTNESS
RGB GAMMA
GRAPHICS
ALPHA
X -Y
COLORIMETRY
SCALING
CORRECT
256 x 32 LUT
CONTRAST SATURATION HUE SHARPNESS
GRAPHICS ONLY
PRIMARY VIDEO SOURCE
BRIGHTNESS
VIDEO PLANE
DECRYPTION
X -Y
COLORIMETRY
SCALING
CORRECT
PRIORITY 4
CONTRAST SATURATION HUE SHARPNESS
RECORDING COLORIMETRY CORRECT
TO OUTPUT FORMATTING
Application Example
OPTIONAL X - Y
BRIGHTNESS
DOWNSCALE
CONTRAST
AND
SATURATION
PROGRESSIVE
HUE
TO INTERLACED
SHARPNESS
OPTIONAL
BRIGHTNESS
FROM
CONSTRAINED
CONTRAST
ALPHA
IMAGE
SATURATION
MIXER
X -Y
HUE
FILTER
SHARPNESS
NTSC / PAL NTSC / PAL ENCODER
COLOR SPACE
VIDEO
CONVERT
DACS
S-VIDEO
YPBPR
BRIGHTNESS CONTRAST SATURATION HUE
COLOR SPACE
HDMI
CONVERT
XMTR
HDMI
SHARPNESS
FROM RECORDING COLORIMETRY CORRECT
OPTIONAL X - Y DOWNSCALE AND PROGRESSIVE
NTSC / PAL ENCODER
TO INTERLACED
Figure 7.63. Video Output Port Processing.
NTSC / PAL
S-VIDEO
263
264
Chapter 7: Digital Video Processing
References 1. Clarke, C. K. P., 1989, Digital Video: Studio Signal Processing, BBC Research Department Report BBC RD1989/14. 2. Devereux, V. G., 1984, Filtering of the Colour-Difference Signals in 4:2:2 YUV Digital Video Coding Systems, BBC Research Department Report BBC RD1984/4. 3. ITU-R BT.601–5, 1995, Studio Encoding Parameters of Digital Television for Standard 4:3 and Widescreen 16:9 Aspect Ratios. 4. ITU-R BT.709–5, 2002, Parameter Values for the HDTV Standards for Production and International Programme Exchange. 5 ITU-R BT.1358, 1998, Studio Parameters of 625 and 525 Line Progressive Scan Television Systems. 6. Johan G.W.M. Janssen, Jeroen H. Stessen, and Peter H.N. de With, An Advanced Sampling Rate Conversion Technique for Video and Graphics Signals, Philips Research Labs. 7. Sandbank, C. P., Digital Television, John Wiley & Sons, Ltd., New York, 1990. 8. SMPTE 274M–2003, Television—1920 × 1080 Image Sample Structure, Digital Representation and Digital Timing Reference Sequences for Multiple Picture Rates.
9. SMPTE 293M–2003, Television—720 × 483 Active Line at 59.94-Hz Progressive Scan Production—Digital Representation. 10. SMPTE 296M–2001, Television—1280 × 720 Progressive Image Sample Structure, Analog and Digital Representation and Analog Interface. 11. SMPTE EG36–1999, Transformations Between Television Component Color Signals. 12. Thomas, G. A., 1996, A Comparison of Motion-Compensated Interlace-to-Progressive Conversion Methods, BBC Research Department Report BBC RD1996/9. 13. Ultimatte®, Technical Bulletin No. 5, Ultimatte Corporation. 14. Watkinson, John, The Engineer’s Guide to Standards Conversion, Snell and Wilcox Handbook Series. 15. Watkinson, John, The Engineer’s Guide to Motion Compensation, Snell and Wilcox Handbook Series.
NTSC Overview
Chapter 8: NTSC, PAL, and SECAM Overview
265
Chapter 8
NTSC, PAL, and SECAM Overview To fully understand the NTSC, PAL, and SECAM encoding and decoding processes, it is helpful to review the background of these standards and how they came about.
NTSC Overview The first color television system was developed in the United States, and on December 17, 1953, the Federal Communications Commission (FCC) approved the transmission standard, with broadcasting approved to begin January 23, 1954. Most of the work for developing a color transmission standard that was compatible with the (then current) 525-line, 60field-per-second, 2:1 interlaced monochrome standard was done by the National Television System Committee (NTSC).
Luminance Information The monochrome luminance (Y) signal is derived from gamma-corrected red, green, and blue (R´G´B´) signals: Y = 0.299R´ + 0.587G´ + 0.114B´
Due to the sound subcarrier at 4.5 MHz, a requirement was made that the color signal fit within the same bandwidth as the monochrome video signal (0–4.2 MHz). For economic reasons, another requirement was made that monochrome receivers must be able to display the black and white portion of a color broadcast and that color receivers must be able to display a monochrome broadcast.
Color Information The eye is most sensitive to spatial and temporal variations in luminance; therefore, luminance information was still allowed the entire bandwidth available (0–4.2 MHz). Color information, to which the eye is less sensitive and which therefore requires less bandwidth, is represented as hue and saturation information. The hue and saturation information is transmitted using a 3.58-MHz subcarrier, encoded so that the receiver can separate the hue, saturation, and luminance information and convert them back to RGB signals for display. Although this allows the transmission of
265
266
Chapter 8: NTSC, PAL, and SECAM Overview
color signals within the same bandwidth as monochrome signals, the problem still remains as to how to separate the color and luminance information cost-effectively, since they occupy the same portion of the frequency spectrum. To transmit color information, U and V or I and Q “color difference” signals are used: R´ – Y = 0.701R´ – 0.587G´ – 0.114B´ B´ – Y = –0.299R´ – 0.587G´ + 0.886B´ U = 0.492(B´ – Y) V = 0.877(R´ – Y) I = 0.596R´ – 0.275G´ – 0.321B´ = Vcos 33° – Usin 33° = 0.736(R´ – Y) – 0.268(B´ – Y) Q = 0.212R´ – 0.523G´ + 0.311B´ = Vsin 33° + Ucos 33° = 0.478(R´ – Y) + 0.413(B´ – Y)
The scaling factors to generate U and V from (B´ – Y) and (R´ – Y) were derived due to overmodulation considerations during transmission. If the full range of (B´ – Y) and (R´ – Y) were used, the modulated chrominance levels would exceed what the monochrome transmitters were capable of supporting. Experimentation determined that modulated subcarrier amplitudes of 20% of the Y signal amplitude could be permitted above white and below black. The scaling factors were then selected so that the maximum level of 75% color would be at the white level. I and Q were initially selected since they more closely related to the variation of color acuity than U and V. The color response of the eye decreases as the size of viewed objects decreases. Small objects, occupying frequencies of 1.3–2.0 MHz, provide little color sensation. Medium objects, occupying the 0.6–1.3 MHz frequency range, are acceptable if repro-
duced along the orange-cyan axis. Larger objects, occupying the 0–0.6 MHz frequency range, require full three-color reproduction. The I and Q bandwidths were chosen accordingly, and the preferred color reproduction axis was obtained by rotating the U and V axes by 33°. The Q component, representing the green-purple color axis, was band-limited to about 0.6 MHz. The I component, representing the orange-cyan color axis, was band-limited to about 1.3 MHz. Another advantage of limiting the I and Q bandwidths to 1.3 MHz and 0.6 MHz, respectively, is to minimize crosstalk due to asymmetrical sidebands as a result of lowpass filtering the composite video signal to about 4.2 MHz. Q is a double sideband signal; however, I is asymmetrical, bringing up the possibility of crosstalk between I and Q. The symmetry of Q avoids crosstalk into I; since Q is bandwidth limited to 0.6 MHz, I crosstalk falls outside the Q bandwidth. U and V, both bandwidth-limited to 1.3 MHz, are now commonly used instead of I and Q. When broadcast, UV crosstalk occurs above 0.6 MHz, however, this is not usually visible due to the limited UV bandwidths used by NTSC decoders for consumer equipment. The UV and IQ vector diagram is shown in Figure 8.1.
Color Modulation I and Q (or U and V) are used to modulate a 3.58-MHz color subcarrier using two balanced modulators operating in phase quadrature: one modulator is driven by the subcarrier at sine phase, the other modulator is driven by the subcarrier at cosine phase. The outputs of the modulators are added together to form the modulated chrominance signal:
NTSC Overview
C = Q sin (ωt + 33°) + I cos (ωt + 33°)
Composite Video Generation
ω = 2πFSC
The modulated chrominance is added to the luminance information along with appropriate horizontal and vertical sync signals, blanking information, and color burst information, to generate the composite color video waveform shown in Figure 8.2.
FSC = 3.579545 MHz (± 10 Hz)
or, if U and V are used instead of I and Q: C = U sin ωt + V cos ωt
Hue information is conveyed by the chrominance phase relative to the subcarrier. Saturation information is conveyed by chrominance amplitude. In addition, if an object has no color (such as a white, gray, or black object), the subcarrier is suppressed.
RED 103˚
composite NTSC = Y + Q sin (ωt + 33°) + I cos (ωt + 33°) + timing
or, if U and V are used instead of I and Q: composite NTSC = Y + U sin ωt + V cos ωt + timing
+V 90˚
IRE SCALE UNITS
MAGENTA 61˚
100 88
80 +Q 33˚
82 60 40 YELLOW 167˚
20
62
BURST 180˚
+U 0˚ 62
82
GREEN 241˚
267
BLUE 347˚
88 –I 303˚ CYAN 283˚
Figure 8.1. UV and IQ Vector Diagram for 75% Color Bars.
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
Chapter 8: NTSC, PAL, and SECAM Overview
WHITE
268
WHITE LEVEL
100 IRE
3.58 MHZ COLOR BURST (9 ± 1 CYCLES)
20 IRE
BLACK LEVEL
7.5 IRE
BLANK LEVEL
20 IRE
40 IRE SYNC LEVEL
BLANK LEVEL
COLOR SATURATION
LUMINANCE LEVEL
PHASE = HUE
Figure 8.2. (M) NTSC Composite Video Signal for 75% Color Bars.
NTSC Overview
The bandwidth of the resulting composite video signal is shown in Figure 8.3. The I and Q (or U and V) information can be transmitted without loss of identity as long as the proper color subcarrier phase relationship is maintained at the encoding and decoding process. A color burst signal, consisting of nine cycles of the subcarrier frequency at a specific phase, follows most horizontal sync pulses, and provides the decoder a reference signal so as to be able to recover the I and Q (or U and V) signals properly. The color burst phase is defined to be along the –U axis as shown in Figure 8.1.
Color Subcarrier Frequency The specific choice for the color subcarrier frequency was dictated by several factors. The first was the need to provide horizontal interlace to reduce the visibility of the subcarrier, requiring that the subcarrier frequency, FSC, be an odd multiple of one-half the horizontal line rate. The second factor was selection of a frequency high enough that it generated a fine interference pattern having low visibility. Third, double sidebands for I and Q (or U and V) bandwidths below 0.6 MHz had to be allowed. The choice of the frequencies is: FH = (4.5 × 106/286) Hz = 15,734.27 Hz FV = FH/(525/2) = 59.94 Hz
FSC = ((13 × 7 × 5)/2) × FH = (455/2) × FH = 3.579545 MHz
The resulting FV (field) and FH (line) rates were slightly different from the monochrome standards, but fell well within the tolerance ranges and were therefore acceptable. Figure 8.4 illustrates the resulting spectral interleaving.
269
The luminance (Y) components are modulated due to the horizontal blanking process, resulting in bunches of luminance information spaced at intervals of FH. These signals are further modulated by the vertical blanking process, resulting in luminance frequency components occurring at NFH ± MFV. N has a maximum value of about 277 with a 4.2-MHz bandwidth-limited luminance. Thus, luminance information is limited to areas about integral harmonics of the line frequency (FH), with additional spectral lines offset from NFH by the 29.97-Hz vertical frame rate. The area in the spectrum between luminance groups, occurring at odd multiples of one-half the line frequency, contains minimal spectral energy and is therefore used for the transmission of chrominance information. The harmonics of the color subcarrier are separated from each other by FH since they are odd multiples of one-half FH, providing a half-line offset and resulting in an interlace pattern that moves upward. Four complete fields are required to repeat a specific sample position, as shown in Figure 8.5.
NTSC Standards Figure 8.6 shows the common designations for NTSC systems. The letter “M” refers to the monochrome standard for line and field rates (525/59.94), a video bandwidth of 4.2 MHz, an audio carrier frequency 4.5 MHz above the video carrier frequency, and a RF channel bandwidth of 6 MHz. The “NTSC” refers to the technique to add color information to the monochrome signal. Detailed timing parameters can be found in Table 8.9. “NTSC 4.43” is commonly used for multistandard analog VCRs. The horizontal and vertical timing is the same as (M) NTSC; color
270
Chapter 8: NTSC, PAL, and SECAM Overview
CHROMINANCE SUBCARRIER
AMPLITUDE
Y
I
I Q
I Q FREQUENCY (MHZ)
0.0
1.0
2.0
3.0
3.58
4.2
(A)
CHROMINANCE SUBCARRIER
AMPLITUDE
Y
U V
U V FREQUENCY (MHZ)
0.0
1.0
2.0
3.0
3.58
4.2
(B)
Figure 8.3. Video Bandwidths of Baseband (M) NTSC Video. (a) Using 1.3-MHz I and 0.6-MHz Q signals. (b) Using 1.3-MHz U and V signals.
NTSC Overview
Y
Y
Y I, Q
I, Q
F FH / 2
FH / 2 FH Y
I, Q
Y
I, Q
Y 29.97 HZ SPACING
F 227FH
228FH 227.5FH
229FH 228.5FH
15.734 KHZ
Figure 8.4. Luma and Chroma Frequency Interleave Principle. Note that 227.5FH = FSC.
271
272
Chapter 8: NTSC, PAL, and SECAM Overview
SERRATION PULSES
ANALOG FIELD 1
523
524
525
1
2
3
4
EQUALIZING PULSES
5
6
7
8
9
10
23
EQUALIZING PULSES
BURST PHASE
ANALOG FIELD 2
261
262
263
264
265
266
524
525
1
268
269
270
271
272
285
286
START OF VSYNC
ANALOG FIELD 3
523
267
2
3
4
5
6
7
8
9
10
23
BURST PHASE
ANALOG FIELD 4
261
262
263
264
265
266
267
268
269
270
271
272
285
BURST BEGINS WITH POSITIVE HALF-CYCLE BURST PHASE = 180˚ RELATIVE TO U HSYNC
HSYNC / 2
BURST BEGINS WITH NEGATIVE HALF-CYCLE BURST PHASE = 180˚ RELATIVE TO U H/2
H/2
H/2
Figure 8.5. Four-field (M) NTSC Sequence and Burst Blanking.
H/2
286
NTSC Overview
273
QUADRATURE MODULATED SUBCARRIER PHASE = HUE AMPLITUDE = SATURATION
"M"
"NTSC–J"
"NTSC 4.43"
LINE / FIELD = 525 / 59.94 FH = 15.734 KHZ FV = 59.94 HZ FSC = 3.579545 MHZ
LINE / FIELD = 525 / 59.94 FH = 15.734 KHZ FV = 59.94 HZ FSC = 3.579545 MHZ
LINE / FIELD = 525 / 59.94 FH = 15.734 KHZ FV = 59.94 HZ FSC = 4.43361875 MHZ
BLANKING SETUP = 7.5 IRE VIDEO BANDWIDTH = 4.2 MHZ AUDIO CARRIER = 4.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 4.2 MHZ AUDIO CARRIER = 4.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
BLANKING SETUP = 7.5 IRE VIDEO BANDWIDTH = 4.2 MHZ AUDIO CARRIER = 4.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
Figure 8.6. Common NTSC Systems.
encoding uses the PAL modulation format and a 4.43361875 MHz color subcarrier frequency. “NTSC–J,” used in Japan, is the same as (M) NTSC, except there is no blanking pedestal during active video. Thus, active video has a nominal amplitude of 714 mV. “Noninterlaced NTSC” is a 262-line, 60 frames-per-second version of NTSC, as shown in Figure 8.7. This format is identical to standard (M) NTSC, except that there are 262 lines per frame.
RF Modulation Figures 8.8, 8.9, and 8.10 illustrate the basic process of converting baseband (M) NTSC composite video to a RF (radio frequency) signal. Figure 8.8a shows the frequency spectrum of a baseband composite video signal. It is similar to Figure 8.3. However, Figure 8.3 only shows the upper sideband for simplicity. The “video carrier” notation at 0 MHz serves only
as a reference point for comparison with Figure 8.8b. Figure 8.8b shows the audio/video signal as it resides within a 6-MHz channel (such as channel 3). The video signal has been lowpass filtered, most of the lower sideband has been removed, and audio information has been added. Figure 8.8c details the information present on the audio subcarrier for stereo (BTSC) operation. As shown in Figures 8.9 and 8.10, back porch clamping (see glossary) of the analog video signal ensures that the back porch level is constant, regardless of changes in the average picture level. White clipping of the video signal prevents the modulated signal from going below 10%; below 10% may result in overmodulation and “buzzing” in television receivers. The video signal is then lowpass filtered to 4.2 MHz and drives the AM (amplitude modulation) video modulator. The sync level corresponds to 100% modulation, the blanking
274
Chapter 8: NTSC, PAL, and SECAM Overview
START OF VSYNC
261
262
1
2
3
4
5
6
7
8
9
10
23
BURST BEGINS WITH POSITIVE HALF-CYCLE BURST PHASE = REFERENCE PHASE = 180˚ RELATIVE TO U
BURST BEGINS WITH NEGATIVE HALF-CYCLE BURST PHASE = REFERENCE PHASE = 180˚ RELATIVE TO U
Figure 8.7. Noninterlaced NTSC Frame Sequence.
corresponds to 75%, and the white level corresponds to 10%. (M) NTSC systems use an IF (intermediate frequency) for the video of 45.75 MHz. At this point, audio information is added on a subcarrier at 41.25 MHz. A monaural audio signal is processed as shown in Figure 8.9 and drives the FM (frequency modulation) modulator. The output of the FM modulator is added to the IF video signal. The SAW filter, used as a vestigial sideband filter, provides filtering of the IF signal. The mixer, or up converter, mixes the IF signal with the desired broadcast frequency. Both sum and difference frequencies are generated by the mixing process, so the difference signal is extracted by using a bandpass filter. Stereo Audio (Analog) BTSC This standard, defined by EIA TVSB5 and known as the BTSC system (Broadcast Television Systems Committee), is shown in Figure 8.10. Countries that use this system include
the United States, Canada, Mexico, Brazil, and Taiwan. To enable stereo, L–R information is transmitted using a suppressed AM subcarrier. A SAP (secondary audio program) channel may also be present, used to transmit a second language or video description (descriptive audio for the visually impaired). A professional channel may also be present, allowing communication with remote equipment and people. Zweiton M This standard (ITU-R BS.707), also known as A2 M, is similar to that used with PAL. The L+R information is transmitted on a FM subcarrier at 4.5 MHz. The L–R information, or a second L+R audio signal, is transmitted on a second FM subcarrier at 4.724212 MHz. If stereo or dual mono signals are present, the FM subcarrier at 4.724212 MHz is amplitude-modulated with a 55.0699 kHz subcarrier. This 55.0699 kHz subcarrier is 50% amplitudemodulated at 149.9 Hz to indicate stereo audio or 276.0 Hz to indicate dual mono audio. This system is used in South Korea.
NTSC Overview
275
VIDEO CARRIER CHROMINANCE SUBCARRIER
CHROMINANCE SUBCARRIER
FREQUENCY (MHZ) –4.5
–4.2
–3.58 –3.0
–1.0
1.0
0.0
3.0
3.58
4.2
4.5
(A)
VIDEO CARRIER
CHROMINANCE SUBCARRIER
0.75 MHZ VESTIGIAL SIDEBAND
AUDIO CARRIER
FREQUENCY (MHZ)
–4.0
–3.0
–0.75
1.0
0.0
3.0
3.58
4.2
4.5
5.0
6 MHZ CHANNEL
–1.25
4.75
(B)
AUDIO CARRIER
FH = 15,734 HZ STEREO PILOT
L + R (FM)
PROFESSIONAL CHANNEL (FM) SAP (FM)
L – R (AM)
FREQUENCY 0.0
FH
2 FH
3 FH
4 FH
5 FH
6.5 FH
(C)
Figure 8.8. Transmission Channel for (M) NTSC. (a) Frequency spectrum of baseband composite video. (b) Frequency spectrum of typical channel including audio information. (c) Detailed frequency spectrum of BTSC stereo audio information.
276
Chapter 8: NTSC, PAL, and SECAM Overview
EIA-J This standard is similar to BTSC, and is used in Japan. The L+R information is transmitted on a FM subcarrier at 4.5 MHz. The L–R signal, or a second L+R signal, is transmitted on a second FM subcarrier at +2FH. If stereo or dual mono signals are present, a +3.5FH subcarrier is amplitude-modulated with either a 982.5 Hz subcarrier (stereo audio) or a 922.5 Hz subcarrier (dual mono audio).
AUDIO LEFT AUDIO RIGHT
L+R --------------75 µS PRE-EMPHASIS --------------50–15000 HZ BPF
Analog Channel Assignments Tables 8.1 through 8.4 list the typical channel assignments for VHF, UHF and cable for various NTSC systems. Note that cable systems routinely reassign channel numbers to alternate frequencies to minimize interference and provide multiple levels of programming (such as regular and preview premium movie channels).
FM MODULATOR
41–47 MHZ BANDWIDTH
(M) NTSC COMPOSITE VIDEO
5 KHZ CLOCK
BACK PORCH CLAMP AND WHITE LEVEL CLIP
4.2 MHZ LPF
AM MODULATOR
+
SAW FILTER
MIXER (UP CONVERTER)
BANDPASS FILTER
MODULATED RF AUDIO / VIDEO (6 MHZ BANDWIDTH)
45.75 MHZ IF VIDEO CARRIER PLL
41.25 MHZ IF AUDIO CARRIER PLL
VIDEO CARRIER OF DESIRED CHANNEL PLL
Figure 8.9. Typical RF Modulation Implementation for (M) NTSC: Mono Audio.
NTSC Overview
150 µS PRE-EMPHASIS --------------300–3,400 HZ BPF
PROFESSIONAL CHANNEL AUDIO
FM MODULATOR
41.25 MHZ – 6.5FH IF AUDIO CARRIER
PROFESSIONAL CHANNEL
50–10,000 HZ BPF --------------BTSC COMPRESSION
SECONDARY AUDIO
FM MODULATOR
SECONDARY AUDIO PROGRAM (SAP)
FM STEREO PILOT SIGNAL
277
+
41.25 MHZ – 5FH IF AUDIO CARRIER
41.25 MHZ – FH
+
L–R --------------50–15,000 HZ BPF --------------BTSC COMPRESSION STEREO MODULATOR
+
L+R --------------75 µS PRE-EMPHASIS --------------50–15,000 HZ BPF
AUDIO LEFT AUDIO RIGHT
41–47 MHZ BANDWIDTH
(M) NTSC COMPOSITE VIDEO
BACK PORCH CLAMP AND WHITE LEVEL CLIP
4.2 MHZ LPF
AM MODULATOR
45.75 MHZ IF VIDEO CARRIER
+
SAW FILTER
MIXER (UP CONVERTER)
BANDPASS FILTER
MODULATED RF AUDIO / VIDEO
CHANNEL SELECT
Figure 8.10. Typical RF Modulation Implementation for (M) NTSC: BTSC Stereo Audio.
278
Chapter 8: NTSC, PAL, and SECAM Overview
Broadcast Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Broadcast Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
– – 2 3 4 5 6 7 8 9
– – 55.25 61.25 67.25 77.25 83.25 175.25 181.25 187.25
– – 59.75 65.75 71.75 81.75 87.75 179.75 185.75 191.75
– – 54–60 60–66 66–72 76–82 82–88 174–180 180–186 186–192
40 41 42 43 44 45 46 47 48 49
627.25 633.25 639.25 645.25 651.25 657.25 663.25 669.25 675.25 681.25
631.75 637.75 643.75 649.75 655.75 661.75 667.75 673.75 679.75 685.75
626–632 632–638 638–644 644–650 650–656 656–662 662–668 668–674 674–680 680–686
10 11 12 13 14 15 16 17 18 19
193.25 199.25 205.25 211.25 471.25 477.25 483.25 489.25 495.25 501.25
197.75 203.75 209.75 215.75 475.75 481.75 487.75 493.75 499.75 505.75
192–198 198–204 204–210 210–216 470–476 476–482 482–488 488–494 494–500 500–506
50 51 52 53 54 55 56 57 58 59
687.25 693.25 699.25 705.25 711.25 717.25 723.25 729.25 735.25 741.25
691.75 697.75 703.75 709.75 715.75 721.75 727.75 733.75 739.75 745.75
686–692 692–698 698–704 704–710 710–716 716–722 722–728 728–734 734–740 740–746
20 21 22 23 24 25 26 27 28 29
507.25 513.25 519.25 525.25 531.25 537.25 543.25 549.25 555.25 561.25
511.75 517.75 523.75 529.75 535.75 541.75 547.75 553.75 559.75 565.75
506–512 512–518 518–524 524–530 530–536 536–542 542–548 548–554 554–560 560–566
60 61 62 63 64 65 66 67 68 69
747.25 753.25 759.25 765.25 771.25 777.25 783.25 789.25 795.25 801.25
751.75 757.75 763.75 769.75 775.75 781.75 787.75 793.75 799.75 805.75
746–752 752–758 758–764 764–770 770–776 776–782 782–788 788–794 794–800 800–806
30 31 32 33 34 35 36 37 38 39
567.25 573.25 579.25 585.25 591.25 597.25 603.25 609.25 615.25 621.25
571.75 577.75 583.75 589.75 595.75 601.75 607.75 613.75 619.75 625.75
566–572 572–578 578–584 584–590 590–596 596–602 602–608 608–614 614–620 620–626
Table 8.1. Analog Broadcast Nominal Frequencies for North America.
NTSC Overview
Broadcast Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Broadcast Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
– 1 2 3 4 5 6 7 8 9
– 91.25 97.25 103.25 171.25 177.25 183.25 189.25 193.25 199.25
– 95.75 101.75 107.75 175.75 181.75 187.75 193.75 197.75 203.75
– 90–96 96–102 102–108 170–176 176–182 182–188 188–194 192–198 198–204
40 41 42 43 44 45 46 47 48 49
633.25 639.25 645.25 651.25 657.25 663.25 669.25 675.25 681.25 687.25
637.75 643.75 649.75 655.75 661.75 667.75 673.75 679.75 685.75 691.75
632–638 638–644 644–650 650–656 656–662 662–668 668–674 674–680 680–686 686–692
10 11 12 13 14 15 16 17 18 19
205.25 211.25 217.25 471.25 477.25 483.25 489.25 495.25 501.25 507.25
209.75 215.75 221.75 475.75 481.75 487.75 493.75 499.75 505.75 511.75
204–210 210–216 216–222 470–476 476–482 482–488 488–494 494–500 500–506 506–512
50 51 52 53 54 55 56 57 58 59
693.25 699.25 705.25 711.25 717.25 723.25 729.25 735.25 741.25 747.25
697.75 703.75 709.75 715.75 721.75 727.75 733.75 739.75 745.75 751.75
692–698 698–704 704–710 710–716 716–722 722–728 728–734 734–740 740–746 746–752
20 21 22 23 24 25 26 27 28 29
513.25 519.25 525.25 531.25 537.25 543.25 549.25 555.25 561.25 567.25
517.75 523.75 529.75 535.75 541.75 547.75 553.75 559.75 565.75 571.75
512–518 518–524 524–530 530–536 536–542 542–548 548–554 554–560 560–566 566–572
60 61 62 – – – – – – –
753.25 759.25 765.25 – – – – – – –
757.75 763.75 769.75 – – – – – – –
752–758 758–764 764–770 – – – – – – –
30 31 32 33 34 35 36 37 38 39
573.25 579.25 585.25 591.25 597.25 603.25 609.25 615.25 621.25 627.25
577.75 583.75 589.75 595.75 601.75 607.75 613.75 619.75 625.75 631.75
572–578 578–584 584–590 590–596 596–602 602–608 608–614 614–620 620–626 626–632
Table 8.2. Analog Broadcast Nominal Frequencies for Japan.
279
280
Chapter 8: NTSC, PAL, and SECAM Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
– – 2 3 4 5 6 7 8 9
– – 55.25 61.25 67.25 77.25 83.25 175.25 181.25 187.25
– – 59.75 65.75 71.75 81.75 87.75 179.75 185.75 191.75
– – 54–60 60–66 66–72 76–82 82–88 174–180 180–186 186–192
40 41 42 43 44 45 46 47 48 49
319.2625 325.2625 331.2750 337.2625 343.2625 349.2625 355.2625 361.2625 367.2625 373.2625
323.7625 329.7625 335.7750 341.7625 347.7625 353.7625 359.7625 365.7625 371.7625 377.7625
318–324 324–330 330–336 336–342 342–348 348–354 354–360 360–366 366–372 372–378
10 11 12 13 14 15 16 17 18 19
193.25 199.25 205.25 211.25 121.2625 127.2625 133.2625 139.25 145.25 151.25
197.75 203.75 209.75 215.75 125.7625 131.7625 137.7625 143.75 149.75 155.75
192–198 198–204 204–210 210–216 120–126 126–132 132–138 138–144 144–150 150–156
50 51 52 53 54 55 56 57 58 59
379.2625 385.2625 391.2625 397.2625 403.25 409.25 415.25 421.25 427.25 433.25
383.7625 389.7625 395.7625 401.7625 407.75 413.75 419.75 425.75 431.75 437.75
378–384 384–390 390–396 396–402 402–408 408–414 414–420 420–426 426–432 432–438
20 21 22 23 24 25 26 27 28 29
157.25 163.25 169.25 217.25 223.25 229.2625 235.2625 241.2625 247.2625 253.2625
161.75 167.75 173.75 221.75 227.75 233.7625 239.7625 245.7625 251.7625 257.7625
156–162 162–168 168–174 216–222 222–228 228–234 234–240 240–246 246–252 252–258
60 61 62 63 64 65 66 67 68 69
439.25 445.55 451.25 457.25 463.25 469.25 475.25 481.25 487.25 493.25
443.75 449.75 455.75 461.75 467.75 473.75 479.75 485.75 491.75 497.75
438–444 444–450 450–456 456–462 462–468 468–474 474–480 480–486 486–492 492–498
30 31 32 33 34 35 36 37 38 39
259.2625 265.2625 271.2625 277.2625 283.2625 289.2625 295.2625 301.2625 307.2625 313.2625
263.7625 269.7625 275.7625 281.7625 287.7625 293.7625 299.7625 305.7625 311.7625 317.7625
258–264 264–270 270–276 276–282 282–288 288–294 294–300 300–306 306–312 312–318
70 71 72 73 74 75 76 77 78 79
499.25 505.25 511.25 517.25 523.25 529.25 535.25 541.25 547.25 553.25
503.75 509.75 515.75 521.75 527.75 533.75 539.75 545.75 551.75 557.75
498–504 504–510 510–516 516–522 522–528 528–534 534–540 540–546 546–552 552–558
Table 8.3a. Standard Analog Cable TV Nominal Frequencies for USA.
NTSC Overview
281
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
80 81 82 83 84 85 86 87 88 89
559.25 565.25 571.25 577.25 583.25 589.25 595.25 601.25 607.25 613.25
563.75 569.75 575.75 581.75 587.75 593.75 599.75 605.75 611.75 617.75
558–564 564–570 570–576 576–582 582–588 588–594 594–600 600–606 606–612 612–618
120 121 122 123 124 125 126 127 128 129
769.25 775.25 781.25 787.25 793.25 799.25 805.25 811.25 817.25 823.25
773.75 779.75 785.75 791.75 797.75 803.75 809.75 815.75 821.75 827.75
768–774 774–780 780–786 786–792 792–798 798–804 804–810 810–816 816–822 822–828
90 91 92 93 94 95 96 97 98 99
619.25 625.25 631.25 637.25 643.25 91.25 97.25 103.25 109.2750 115.2750
623.75 629.75 635.75 641.75 647.75 95.75 101.75 107.75 113.7750 119.7750
618–624 624–630 630–636 636–642 642–648 90–96 96–102 102–108 108–114 114–120
130 131 132 133 134 135 136 137 138 139
829.25 835.25 841.25 847.25 853.25 859.25 865.25 871.25 877.25 883.25
833.75 839.75 845.75 851.75 857.75 863.75 869.75 875.75 881.75 887.75
828–834 834–840 840–846 846–852 852–858 858–864 864–870 870–876 876–882 882–888
100 101 102 103 104 105 106 107 108 109
649.25 655.25 661.25 667.25 673.25 679.25 685.25 691.25 697.25 703.25
653.75 659.75 665.75 671.75 677.75 683.75 689.75 695.75 701.75 707.75
648–654 654–660 660–666 666–672 672–678 678–684 684–690 690–696 696–702 702–708
140 141 142 143 144 145 146 147 148 149
889.25 895.25 901.25 907.25 913.25 919.25 925.25 931.25 937.25 943.25
893.75 899.75 905.75 911.75 917.75 923.75 929.75 935.75 941.75 947.75
888–894 894–900 900–906 906–912 912–918 918–924 924–930 930–936 936–942 942–948
110 111 112 113 114 115 116 117 118 119
709.25 715.25 721.25 727.25 733.25 739.25 745.25 751.25 757.25 763.25
713.75 719.75 725.75 731.75 737.75 743.75 749.75 755.75 761.75 767.75
708–714 714–720 720–726 726–732 732–738 738–744 744–750 750–756 756–762 762–768
150 151 152 153 154 155 156 157 158 –
949.25 955.25 961.25 967.25 973.25 979.25 985.25 991.25 997.25 –
953.75 959.75 965.75 971.75 977.75 983.75 989.75 995.75 1001.75 –
948–954 954–960 960–966 966–972 972–978 978–984 984–990 990–996 996–1002 –
Table 8.3b. Standard Analog Cable TV Nominal Frequencies for USA.
282
Chapter 8: NTSC, PAL, and SECAM Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
– 1 2 3 4 5 6 7 8 9
– 73.2625 55.2625 61.2625 67.2625 79.2625 85.2625 175.2625 181.2625 187.2625
– 77.7625 59.7625 65.7625 71.7625 83.7625 89.7625 179.7625 185.7625 191.7625
40 41 42 43 44 45 46 47 48 49
319.2625 325.2625 331.2750 337.2625 343.2625 349.2625 355.2625 361.2625 367.2625 373.2625
323.7625 329.7625 335.7750 341.7625 347.7625 353.7625 359.7625 365.7625 371.7625 377.7625
10 11 12 13 14 15 16 17 18 19
193.2625 199.2625 205.2625 211.2625 121.2625 127.2625 133.2625 139.2625 145.2625 151.2625
197.7625 203.7625 209.7625 215.7625 125.7625 131.7625 137.7625 143.7625 149.7625 155.7625
50 51 52 53 54 55 56 57 58 59
379.2625 385.2625 391.2625 397.2625 403.2625 409.2625 415.2625 421.2625 427.2625 433.2625
383.7625 389.7625 395.7625 401.7625 407.7625 413.7625 419.7625 425.7625 431.7625 437.7625
20 21 22 23 24 25 26 27 28 29
157.2625 163.2625 169.2625 217.2625 223.2625 229.2625 235.2625 241.2625 247.2625 253.2625
161.7625 167.7625 173.7625 221.7625 227.7625 233.7625 239.7625 245.7625 251.7625 257.7625
60 61 62 63 64 65 66 67 68 69
439.2625 445.2625 451.2625 457.2625 463.2625 469.2625 475.2625 481.2625 487.2625 493.2625
443.7625 449.7625 455.7625 461.7625 467.7625 473.7625 479.7625 485.7625 491.7625 497.7625
30 31 32 33 34 35 36 37 38 39
259.2625 265.2625 271.2625 277.2625 283.2625 289.2625 295.2625 301.2625 307.2625 313.2625
263.7625 269.7625 275.7625 281.7625 287.7625 293.7625 299.7625 305.7625 311.7625 317.7625
70 71 72 73 74 75 76 77 78 79
499.2625 505.2625 511.2625 517.2625 523.2625 529.2625 535.2625 541.2625 547.2625 553.2625
503.7625 509.7625 515.7625 521.7625 527.7625 533.7625 539.7625 545.7625 551.7625 557.7625
Table 8.3c. Analog Cable TV Nominal Frequencies for USA: Incrementally Related Carrier (IRC) Systems.
NTSC Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
80 81 82 83 84 85 86 87 88 89
559.2625 565.2625 571.2625 577.2625 583.2625 589.2625 595.2625 601.2625 607.2625 613.2625
563.7625 569.7625 575.7625 581.7625 587.7625 593.7625 599.7625 605.7625 611.7625 617.7625
120 121 122 123 124 125 126 127 128 129
769.2625 775.2625 781.2625 787.2625 793.2625 799.2625 805.2625 811.2625 817.2625 823.2625
773.7625 779.7625 785.7625 791.7625 797.7625 803.7625 809.7625 815.7625 821.7625 827.7625
90 91 92 93 94 95 96 97 98 99
619.2625 625.2625 631.2625 637.2625 643.2625 91.2625 97.2625 103.2625 109.2750 115.2625
623.7625 629.7625 635.7625 641.7625 647.7625 95.7625 101.7625 107.7625 113.7750 119.7625
130 131 132 133 134 135 136 137 138 139
829.2625 835.2625 841.2625 847.2625 853.2625 859.2625 865.2625 871.2625 877.2625 883.2625
833.7625 839.7625 845.7625 851.7625 857.7625 863.7625 869.7625 875.7625 881.7625 887.7625
100 101 102 103 104 105 106 107 108 109
649.2625 655.2625 661.2625 667.2625 673.2625 679.2625 685.2625 691.2625 697.2625 703.2625
653.7625 659.7625 665.7625 671.7625 677.7625 683.7625 689.7625 695.7625 701.7625 707.7625
140 141 142 143 144 145 146 147 148 149
889.2625 895.2625 901.2625 907.2625 913.2625 919.2625 925.2625 931.2625 937.2625 943.2625
893.7625 899.7625 905.7625 911.7625 917.7625 923.7625 929.7625 935.7625 941.7625 947.7625
110 111 112 113 114 115 116 117 118 119
709.2625 715.2625 721.2625 727.2625 733.2625 739.2625 745.2625 751.2625 757.2625 763.2625
713.7625 719.7625 725.7625 731.7625 737.7625 743.7625 749.7625 755.7625 761.7625 767.7625
150 151 152 153 154 155 156 157 158 –
949.2625 955.2625 961.2625 967.2625 973.2625 979.2625 985.2625 991.2625 997.2625 –
953.7625 959.7625 965.7625 971.7625 977.7625 983.7625 989.7625 995.7625 1001.7625 –
Table 8.3d. Analog Cable TV Nominal Frequencies for USA: Incrementally Related Carrier (IRC) Systems.
283
284
Chapter 8: NTSC, PAL, and SECAM Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
– 1 2 3 4 5 6 7 8 9
– 72.0036 54.0027 60.0030 66.0033 72.0036 78.0039 174.0087 180.0090 186.0093
– 76.5036 58.5027 64.5030 70.5030 82.5039 88.5042 178.5087 184.5090 190.5093
40 41 42 43 44 45 46 47 48 49
318.0159 324.0162 330.0165 336.0168 342.0168 348.0168 354.0168 360.0168 366.0168 372.0168
322.5159 328.5162 334.5165 340.5168 346.5168 352.5168 358.5168 364.5168 370.5168 376.5168
10 11 12 13 14 15 16 17 18 19
192.0096 198.0099 204.0102 210.0105 120.0060 126.0063 132.0066 138.0069 144.0072 150.0075
196.5096 202.5099 208.5102 214.5105 124.5060 130.5063 136.5066 142.5069 148.5072 154.5075
50 51 52 53 54 55 56 57 58 59
378.0168 384.0168 390.0168 396.0168 402.0201 408.0204 414.0207 420.0210 426.0213 432.0216
382.5168 388.5168 394.5168 400.5168 406.5201 412.5204 418.5207 424.5210 430.5213 436.5216
20 21 22 23 24 25 26 27 28 29
156.0078 162.0081 168.0084 216.0108 222.0111 228.0114 234.0117 240.0120 246.0123 252.0126
160.5078 166.5081 172.5084 220.5108 226.5111 232.5114 238.5117 244.5120 250.5123 256.5126
60 61 62 63 64 65 66 67 68 69
438.0219 444.0222 450.0225 456.0228 462.0231 468.0234 474.0237 480.0240 486.0243 492.0246
442.5219 448.5222 454.5225 460.5228 466.5231 472.5234 478.5237 484.5240 490.5243 496.5246
30 31 32 33 34 35 36 37 38 39
258.0129 264.0132 270.0135 276.0138 282.0141 288.0144 294.0147 300.0150 306.0153 312.0156
262.5129 268.5132 274.5135 280.5138 286.5141 292.5144 298.5147 304.5150 310.5153 316.5156
70 71 72 73 74 75 76 77 78 79
498.0249 504.0252 510.0255 516.0258 522.0261 528.0264 534.0267 540.0270 546.0273 552.0276
502.5249 508.5252 514.5255 520.5258 526.5261 532.5264 538.5267 544.5270 550.5273 556.5276
Table 8.3e. Analog Cable TV Nominal Frequencies for USA: Harmonically Related Carrier (HRC) systems.
NTSC Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
80 81 82 83 84 85 86 87 88 89
558.0279 564.0282 570.0285 576.0288 582.0291 588.0294 594.0297 600.0300 606.0303 612.0306
562.5279 568.5282 574.5285 580.5288 586.5291 592.5294 598.5297 604.5300 610.5303 616.5306
120 121 122 123 124 125 126 127 128 129
768.0384 774.0387 780.0390 786.0393 792.0396 798.0399 804.0402 810.0405 816.0408 822.0411
772.5384 778.5387 784.5390 790.5393 796.5396 802.5399 808.5402 814.5405 820.5408 826.5411
90 91 92 93 94 95 96 97 98 99
618.0309 624.0312 630.0315 636.0318 642.0321 90.0045 96.0048 102.0051 – –
622.5309 628.5312 634.5315 640.5318 646.5321 94.5045 100.5048 106.5051 – –
130 131 132 133 134 135 136 137 138 139
828.0414 834.0417 840.0420 846.0423 852.0426 858.0429 864.0432 870.0435 876.0438 882.0441
832.5414 838.5417 844.5420 850.5423 856.5426 862.5429 868.5432 874.5435 880.5438 888.5441
100 101 102 103 104 105 106 107 108 109
648.0324 654.0327 660.0330 666.0333 672.0336 678.0339 684.0342 690.0345 696.0348 702.0351
652.5324 658.5327 664.5330 670.5333 676.5336 682.5339 688.5342 694.5345 700.5348 706.5351
140 141 142 143 144 145 146 147 148 149
888.0444 894.0447 900.0450 906.0453 912.0456 918.0459 924.0462 930.0465 936.0468 942.0471
892.5444 898.5447 904.5450 910.5453 916.5456 922.5459 928.5462 934.5465 940.5468 946.5471
110 111 112 113 114 115 116 117 118 119
708.0354 714.0357 720.0360 726.0363 732.0366 738.0369 744.0372 750.0375 756.0378 762.0381
712.5354 718.5357 724.5360 730.5363 736.5366 742.5369 748.5372 754.5375 760.5378 766.5381
150 151 152 153 154 155 156 157 158 –
948.0474 954.0477 960.0480 966.0483 972.0486 978.0489 984.0492 990.0495 996.0498 –
952.5474 958.5477 964.5480 970.5483 976.5486 982.5489 988.5492 994.5495 1000.5498 –
Table 8.3f. Analog Cable TV Nominal Frequencies for USA: Harmonically Related Carrier (HRC) systems.
285
286
Chapter 8: NTSC, PAL, and SECAM Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
– – – 13 14 15 16 17 18 19
– – – 109.25 115.25 121.25 127.25 133.25 139.25 145.25
– – – 113.75 119.75 125.75 131.75 137.75 143.75 149.75
– – – 108–114 114–120 120–126 126–132 132–138 138–144 144–150
40 41 42 46 44 45 46 47 48 49
325.25 331.25 337.25 343.25 349.25 355.25 361.25 367.25 373.25 379.25
329.75 335.75 341.75 347.75 353.75 359.75 365.75 371.75 377.75 383.75
324–330 330–336 336–342 342–348 348–354 354–360 360–366 366–372 372–378 378–384
20 21 22 23 24 25 26 27 28 29
151.25 157.25 165.25 223.25 231.25 237.25 243.25 249.25 253.25 259.25
155.75 161.75 169.75 227.75 235.75 241.75 247.75 253.75 257.75 263.75
150–156 156–162 164–170 222–228 230–236 236–242 242–248 248–254 252–258 258–264
50 51 52 53 54 55 56 57 58 59
385.25 391.25 397.25 403.25 409.25 415.25 421.25 427.25 433.25 439.25
389.75 395.75 401.75 407.75 413.75 419.75 425.75 431.75 437.75 443.75
384–390 390–396 396–402 402–408 408–414 414–420 420–426 426–432 432–438 438–444
30 31 32 33 34 35 36 37 38 39
265.25 271.25 277.25 283.25 289.25 295.25 301.25 307.25 313.25 319.25
269.75 275.75 281.75 287.75 293.75 299.75 305.75 311.75 317.75 323.75
264–270 270–276 276–282 282–288 288–294 294–300 300–306 306–312 312–318 318–324
60 61 62 63 – – – – – –
445.25 451.25 457.25 463.25 – – – – – –
449.75 455.75 461.75 467.75 – – – – – –
444–450 450–456 456–462 462–468 – – – – – –
Table 8.4. Analog Cable TV Nominal Frequencies for Japan.
287
NTSC Overview
Luminance Equation Derivation The equation for generating luminance from RGB is determined by the chromaticities of the three primary colors used by the receiver and what color white actually is. The chromaticities of the RGB primaries and reference white (CIE illuminate C) were specified in the 1953 NTSC standard to be: R:
xr = 0.67
yr = 0.33
zr = 0.00
G:
xg = 0.21
yg = 0.71
zg = 0.08
B:
xb = 0.14
yb = 0.08
zb = 0.78
white: xw = 0.3101 zw = 0.3737
xr xg xb Kr yr yg yb Kg =
Kg = Kb
Kg Kb
0.3101 ⁄ 0.3162 0.67 0.21 0.14 = 1 0.33 0.71 0.08 0.3737 ⁄ 0.3162 0.00 0.08 0.78
=
–1
0.9807 1.730 – 0.482 – 0.261 1 – 0.814 1.652 – 0.023 1.1818 0.083 – 0.169 1.284 0.906
= 0.827
xw ⁄ yw 1 zw ⁄ yw
Rearranging to solve for Kr, Kg, and Kb yields: Kr
Kr
yw = 0.3162
where x and y are the specified CIE 1931 chromaticity coordinates; z is calculated by knowing that x + y + z = 1. Luminance is calculated as a weighted sum of RGB, with the weights representing the actual contributions of each of the RGB primaries in generating the luminance of reference white. We find the linear combination of RGB that gives reference white by solving the equation:
zr zg zb Kb
Substituting the known values gives us the solution for Kr, Kg, and Kb:
xw ⁄ yw xr xg xb yr yg yb 1 zw ⁄ yw zr zg zb
–1
1.430
Y is defined to be Y = (Kr yr)R´ + (Kgyg)G´ + (Kbyb)B´ = (0.906)(0.33)R´ + (0.827)(0.71)G´ + (1.430)(0.08)B´
or Y = 0.299R´ + 0.587G´ + 0.114B´
Modern receivers use a different set of RGB phosphors, resulting in slightly different chromaticities of the RGB primaries and reference white (CIE illuminate D65): R:
xr = 0.630
yr = 0.340
zr = 0.030
G:
xg = 0.310
yg = 0.595
zg = 0.095
B:
xb = 0.155
yb = 0.070
zb = 0.775
white: xw = 0.3127 zw = 0.3583
yw = 0.3290
where x and y are the specified CIE 1931 chromaticity coordinates; z is calculated by knowing that x + y + z = 1. Once again, substituting the known values gives us the solution for Kr, Kg, and Kb:
288
Chapter 8: NTSC, PAL, and SECAM Overview
Kr = Kg = Kb
=
0.3127 ⁄ 0.3290 0.630 0.310 0.155 0.340 0.595 0.070 1 0.3583 ⁄ 0.3290 0.030 0.095 0.775
PAL Overview –1
0.6243 1.1770 1.2362
Since Y is defined to be Y = (Kr yr)R´ + (Kgyg)G´ + (Kbyb)B´ = (0.6243)(0.340)R´ + (1.1770)(0.595)G´ + (1.2362)(0.070)B´
Europe delayed adopting a color television standard, evaluating various systems between 1953 and 1967 that were compatible with their 625-line, 50-field-per-second, 2:1 interlaced monochrome standard. The NTSC specification was modified to overcome the high order of phase and amplitude integrity required during broadcast to avoid color distortion. The Phase Alternation Line (PAL) system implements a line-by-line reversal of the phase of one of the color components, originally relying on the eye to average any color distortions to the correct color. Broadcasting began in 1967 in Germany and the United Kingdom, with each using a slightly different variant of the PAL system.
this results in:
Luminance Information Y = 0.212R´ + 0.700G´ + 0.086B´
However, the standard Y = 0.299R´ + 0.587G´ + 0.114B´ equation is still used. Adjustments are made in the receiver to minimize color errors.
The monochrome luminance (Y) signal is derived from R´G´B´: Y = 0.299R´ + 0.587G´ + 0.114B´
As with NTSC, the luminance signal occupies the entire video bandwidth. PAL has several variations, depending on the video bandwidth and placement of the audio subcarrier. The composite video signal has a bandwidth of 4.2, 5.0, 5.5, or 6.0 MHz, depending on the specific PAL standard.
PAL Overview
Color Information To transmit color information, U and V are used: U = 0.492(B´ – Y) V = 0.877(R´ – Y)
U and V have a typical bandwidth of 1.3 MHz.
Color Modulation As in the NTSC system, U and V are used to modulate the color subcarrier using two balanced modulators operating in phase quadrature: one modulator is driven by the subcarrier at sine phase, the other modulator is driven by the subcarrier at cosine phase. The outputs of the modulators are added together to form the modulated chrominance signal: C = U sin ωt ± V cos ωt ω = 2πFSC
FSC = 4.43361875 MHz (± 5 Hz) for (B, D, G, H, I, N) PAL FSC = 3.58205625 MHz (± 5 Hz) for (NC) PAL FSC = 3.57561149 MHz (± 10 Hz) for (M) PAL
In PAL, the phase of V is reversed every other line. V was chosen for the reversal process since it has a lower gain factor than U and therefore is less susceptible to a one-half FH switching rate imbalance. The result of alternating the V phase at the line rate is that any color subcarrier phase errors produce complementary errors, allowing line-to-line averaging at the receiver to cancel the errors and generate the correct hue with slightly reduced saturation. This technique requires the PAL receiver to be able to determine the correct V phase. This is done using a technique known as AB sync, PAL sync, PAL Switch, or “swing-
289
ing burst,” consisting of alternating the phase of the color burst by ±45° at the line rate. The UV vector diagrams are shown in Figures 8.11 and 8.12. “Simple” PAL decoders rely on the eye to average the line-by-line hue errors. “Standard” PAL decoders use a 1-H delay line to separate U from V in an averaging process. Both implementations have the problem of Hanover bars, in which pairs of adjacent lines have a real and complementary hue error. Chrominance vertical resolution is reduced as a result of the line averaging process.
Composite Video Generation The modulated chrominance is added to the luminance information along with appropriate horizontal and vertical sync signals, blanking signals, and color burst signals, to generate the composite color video waveform shown in Figure 8.13. composite PAL = Y + U sin ωt ± V cos ωt + timing
The bandwidth of the resulting composite video signal is shown in Figure 8.14. Like NTSC, the luminance components are spaced at FH intervals due to horizontal blanking. Since the V component is switched symmetrically at one-half the line rate, only odd harmonics are generated, resulting in V components that are spaced at intervals of FH. The V components are spaced at half-line intervals from the U components, which also have FH spacing. If the subcarrier had a half-line offset like NTSC uses, the U components would be perfectly interleaved, but the V components would coincide with the Y components and thus not be interleaved, creating vertical stationary dot patterns. For this reason, PAL uses a 1/4 line offset for the subcarrier frequency:
290
Chapter 8: NTSC, PAL, and SECAM Overview
IRE SCALE UNITS RED 103˚
+V 90˚ MAGENTA 61˚
100
BURST 135˚
80
95
89 60 40 YELLOW 167˚
20
67
+U 0˚ 67
89
BLUE 347˚
95
GREEN 241˚
CYAN 283˚
Figure 8.11. UV Vector Diagram for 75% Color Bars. Line [n], PAL Switch = zero. +V 90˚
IRE SCALE UNITS CYAN 77˚
GREEN 120˚
100 80
95
89 60 40 20
67
BLUE 13˚
+U 0˚ 67
YELLOW 193˚
89 BURST 225˚
95
MAGENTA 300˚ RED 257˚
Figure 8.12. UV Vector Diagram for 75% Color Bars. Line [n + 1], PAL Switch = one.
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
WHITE
PAL Overview
WHITE LEVEL
100 IRE COLOR BURST (10 ± 1 CYCLES)
21.43 IRE
BLACK / BLANK LEVEL 21.43 IRE
43 IRE SYNC LEVEL
BLANK LEVEL
COLOR SATURATION
LUMINANCE LEVEL
PHASE = HUE
Figure 8.13. (B, D, G, H, I, NC) PAL Composite Video Signal for 75% Color Bars.
291
292
Chapter 8: NTSC, PAL, and SECAM Overview
CHROMINANCE SUBCARRIER
AMPLITUDE
Y
U ±V
U ±V FREQUENCY (MHZ)
0.0
1.0
2.0
3.0
4.0 4.43
5.0
5.5
(I) PAL
CHROMINANCE SUBCARRIER
AMPLITUDE
Y
U ±V
U ±V FREQUENCY (MHZ)
0.0
1.0
2.0
3.0
4.0 4.43
5.0
(B, G, H) PAL
Figure 8.14. Video Bandwidths of Some PAL Systems.
Y U
Y V
U
V
F FH / 4 FH / 2 FH
Figure 8.15. Luma and Chroma Frequency Interleave Principle.
PAL Overview
FSC = ((1135/4) + (1/625)) FH for (B, D, G, H, I, N) PAL FSC = (909/4) FH for (M) PAL FSC = ((917/4) + (1/625)) FH for (NC) PAL
The additional (1/625) FH factor (equal to 25 Hz) provides motion to the color dot pattern, reducing its visibility. Figure 8.15 illustrates the resulting frequency interleaving. Eight complete fields are required to repeat a specific sample position, as shown in Figures 8.16 and 8.17.
PAL Standards Figure 8.19 shows the common designations for PAL systems. The letters refer to the monochrome standard for line and field rate, video bandwidth (4.2, 5.0, 5.5, or 6.0 MHz), audio carrier relative frequency, and RF channel bandwidth (6.0, 7.0, or 8.0 MHz). The “PAL” refers to the technique to add color information to the monochrome signal. Detailed timing parameters may be found in Table 8.9. “Noninterlaced PAL,” shown in Figure 8.18, is a 312-line, 50 frames-per-second version of PAL common among video games and on-screen displays. This format is identical to standard PAL, except that there are 312 lines per frame.
RF Modulation Figures 8.20 and 8.21 illustrate the process of converting baseband (G) PAL composite video to an RF (radio frequency) signal. The process for the other PAL standards is similar, except primarily for the different video bandwidths and subcarrier frequencies.
293
Figure 8.20a shows the frequency spectrum of a (G) PAL baseband composite video signal. It is similar to Figure 8.14. However, Figure 8.14 only shows the upper sideband for simplicity. The “video carrier” notation at 0 MHz serves only as a reference point for comparison with Figure 8.20b. Figure 8.20b shows the audio/video signal as it resides within an 8 MHz channel. The video signal has been lowpass filtered, most of the lower sideband has been removed, and audio information has been added. Note that (H) and (I) PAL have a vestigial sideband of 1.25 MHz, rather than 0.75 MHz. Figure 8.20c details the information present on the audio subcarrier for analog stereo operation. As shown in Figure 8.21, back porch clamping of the analog video signal ensures that the back porch level is constant, regardless of changes in the average picture level. The video signal is then lowpass filtered to 5.0 MHz and drives the AM (amplitude modulation) video modulator. The sync level corresponds to 100% modulation; the blanking and white modulation levels are dependent on the specific version of PAL: blanking level (% modulation) B, G D, H, M, N I
75% 75% 76%
white level (% modulation) B, G, H, M, N D I
10% 10% 20%
Note that PAL systems use a variety of video and audio IF frequencies (values in MHz):
294
Chapter 8: NTSC, PAL, and SECAM Overview START OF VSYNC
ANALOG FIELD 1
620
621
622
623
624
625
1
2
3
4
5
6
7
23
24
–U COMPONENT OF BURST PHASE
ANALOG FIELD 2
308
309
310
311
312
313
314
315
316
317
318
319
320
336
337
ANALOG FIELD 3
620
621
622
623
624
625
1
2
3
4
5
6
7
23
24
ANALOG FIELD 4
308
309
310
311
312
313
314
315
316
317
318
319
320
FIELD ONE BURST BLANKING INTERVALS
FIELD TWO FIELD THREE FIELD FOUR
BURST PHASE = REFERENCE PHASE = 135˚ RELATIVE TO U PAL SWITCH = 0, + V COMPONENT
BURST PHASE = REFERENCE PHASE + 90˚ = 225˚ RELATIVE TO U PAL SWITCH = 1, – V COMPONENT
Figure 8.16a. Eight-field (B, D, G, H, I, NC) PAL Sequence and Burst Blanking. See Figure 8.5 for equalization and serration pulse details.
336
337
295
PAL Overview START OF VSYNC
ANALOG FIELD 5
620
621
622
623
624
625
1
2
3
4
5
6
7
23
24
–U COMPONENT OF BURST PHASE
ANALOG FIELD 6
308
309
310
311
312
313
314
315
316
317
318
319
320
336
337
ANALOG FIELD 7
620
621
622
623
624
625
1
2
3
4
5
6
7
23
24
ANALOG FIELD 8
308
309
310
311
312
313
314
315
316
317
318
319
320
FIELD FIVE BURST BLANKING INTERVALS
FIELD SIX FIELD SEVEN FIELD EIGHT
BURST PHASE = REFERENCE PHASE = 135˚ RELATIVE TO U PAL SWITCH = 0, + V COMPONENT
BURST PHASE = REFERENCE PHASE + 90˚ = 225˚ RELATIVE TO U PAL SWITCH = 1, – V COMPONENT
Figure 8.16b. Eight-field (B, D, G, H, I, NC) PAL Sequence and Burst Blanking. See Figure 8.5 for equalization and serration pulse details.
336
337
296
Chapter 8: NTSC, PAL, and SECAM Overview START OF VSYNC ANALOG FIELD 1 / 5
520
521
522
523
524
525
1
2
3
4
5
6
7
8
9
ANALOG FIELD 2 / 6
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
ANALOG FIELD 3 / 7
520
521
522
523
524
525
1
2
3
4
5
6
7
8
9
ANALOG FIELD 4 / 8
258
259
260
261
262
263
264
265
266
267
268
269
270
BURST PHASE = REFERENCE PHASE = 135˚ RELATIVE TO U PAL SWITCH = 0, + V COMPONENT
BURST PHASE = REFERENCE PHASE + 90˚ = 225˚ RELATIVE TO U PAL SWITCH = 1, – V COMPONENT
Figure 8.17. Eight-field (M) PAL Sequence and Burst Blanking. See Figure 8.5 for equalization and serration pulse details.
271
272
297
PAL Overview
START OF VSYNC
308
309
310
311
312
1
2
3
4
5
6
7
23
24
BURST PHASE = REFERENCE PHASE = 135˚ RELATIVE TO U PAL SWITCH = 0, + V COMPONENT
BURST PHASE = REFERENCE PHASE + 90˚ = 225˚ RELATIVE TO U PAL SWITCH = 1, – V COMPONENT
Figure 8.18. Noninterlaced PAL Frame Sequence.
B, G B D D I I M, N
video
audio
38.900 36.875 37.000 38.900 38.900 39.500 45.750
33.400 31.375 30.500 32.400 32.900 33.500 41.250
Australia China OIRT U.K.
At this point, audio information is added on the audio subcarrier. A monaural L+R audio signal is processed as shown in Figure 8.21 and drives the FM (frequency modulation) modulator. The output of the FM modulator is added to the IF video signal. The SAW filter, used as a vestigial sideband filter, provides filtering of the IF signal. The mixer, or up converter, mixes the IF signal with the desired broadcast frequency. Both sum and difference frequencies are generated by the mixing process, so the difference signal is extracted by using a bandpass filter.
Stereo Audio (Analog) The standard (ITU-R BS.707), also known as Zweiton or A2, is shown in Figure 8.21. The L+R information is transmitted on a FM subcarrier. The R information, or a second L+R audio signal, is transmitted on a second FM subcarrier at +15.5FH. If stereo or dual mono signals are present, the FM subcarrier at +15.5FH is amplitudemodulated with a 54.6875 kHz (3.5FH) subcarrier. This 54.6875 kHz subcarrier is 50% amplitude-modulated at 117.5 Hz (FH / 133) to indicate stereo audio or 274.1 Hz (FH / 57) to indicate dual mono audio. Countries that use this system include Australia, Austria, China, Germany, Italy, Malaysia, Netherlands, Slovenia, and Switzerland. Stereo Audio (Digital) The standard uses NICAM 728 (Near Instantaneous Companded Audio Multiplex),
298
Chapter 8: NTSC, PAL, and SECAM Overview
QUADRATURE MODULATED SUBCARRIER PHASE = HUE AMPLITUDE = SATURATION LINE ALTERNATION OF V COMPONENT
"B, B1, G, H"
"I"
"M"
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ FSC = 4.43361875 MHZ
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ FSC = 4.43361875 MHZ
LINE / FIELD = 525 / 59.94 FH = 15.734 KHZ FV = 59.94 HZ FSC = 3.57561149 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 5.5 MHZ AUDIO CARRIER = 5.9996 MHZ CHANNEL BANDWIDTH = 8 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 5.0 MHZ AUDIO CARRIER = 5.5 MHZ CHANNEL BANDWIDTH: B = 7 MHZ B1, G, H = 8 MHZ
BLANKING SETUP = 7.5 IRE VIDEO BANDWIDTH = 4.2 MHZ AUDIO CARRIER = 4.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
"D"
"N"
"NC"
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ FSC = 4.43361875 MHZ
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ FSC = 4.43361875 MHZ
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ FSC = 3.58205625 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 6.0 MHZ AUDIO CARRIER = 6.5 MHZ CHANNEL BANDWIDTH = 8 MHZ
BLANKING SETUP = 7.5 IRE VIDEO BANDWIDTH = 5.0 MHZ AUDIO CARRIER = 5.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 4.2 MHZ AUDIO CARRIER = 4.5 MHZ CHANNEL BANDWIDTH = 6 MHZ
Figure 8.19. Common PAL Systems.
PAL Overview
299
VIDEO CARRIER CHROMINANCE SUBCARRIER
CHROMINANCE SUBCARRIER
FREQUENCY (MHZ)
–5.5
–5.0
–4.43
0.0
–1.0
1.0
4.43
5.0
5.5
(A)
VIDEO CARRIER
CHROMINANCE SUBCARRIER
0.75 MHZ VESTIGIAL SIDEBAND
AUDIO CARRIER
FREQUENCY (MHZ) –4.0
–3.0
–0.75
0.0
4.43
1.0
5.0
5.5
8 MHZ CHANNEL
–1.25
6.75
(B)
AUDIO CARRIER
FH = 15,625 HZ
R (FM)
L + R (FM)
FREQUENCY
–50 KHZ
0.0
50 KHZ
15.5 FH – 50 KHZ
15.5 FH
15.5 FH + 50 KHZ
(C)
Figure 8.20. Transmission Channel for (G) PAL. (a) Frequency spectrum of baseband composite video. (b) Frequency spectrum of typical channel including audio information. (c) Detailed frequency spectrum of Zweiton analog stereo audio information.
300
Chapter 8: NTSC, PAL, and SECAM Overview
discussed within BS.707 and ETSI EN 300 163. It was developed by the BBC and IBA to increase sound quality, provide multiple channels of digital sound or data, and be more resistant to transmission interference. The subcarrier resides either 5.85 MHz above the video carrier for (B, D, G, H) PAL and (L) SECAM systems or 6.552 MHz above the video carrier for (I) PAL systems. Countries that use NICAM 728 include Belgium, China, Denmark, Finland, France, Hungary, New Zealand, Norway, Singapore, South Africa, Spain, Sweden, and the United Kingdom. NICAM 728 is a digital system that uses a 32 kHz sampling rate and 14-bit resolution. A bit rate of 728 kbps is used, giving it the name
NICAM 728. Data is transmitted in frames, with each frame containing 1 ms of audio. As shown in Figure 8.22, each frame consists of: 8-bit frame alignment word (01001110) 5 control bits (C0–C4) 11 undefined bits (AD0–AD10) 704 audio data bits (A000–A703)
C0 is a “1” for eight successive frames and a “0” for the next eight frames, defining a 16frame sequence. C1–C3 specify the format transmitted: “000” = one stereo signal with the left channel being odd-numbered samples and the right channel being even-numbered samples, “010” = two independent mono channels transmitted in alternate frames, “100” = one mono channel and one 352 kbps data channel 117.5 HZ 274.1 HZ
STEREO PILOT SIGNAL
3.5FH
AM MODULATOR
50 µS PRE-EMPHASIS --------------40–15,000 HZ BPF
FM MODULATOR
AM MODULATOR
33.4 MHZ – 15.5FH IF AUDIO CARRIER
L+R --------------50 µS PRE-EMPHASIS --------------40–15,000 HZ BPF
AUDIO RIGHT AUDIO LEFT
FM MODULATOR
+
33.4 MHZ IF AUDIO CARRIER
(G) PAL COMPOSITE VIDEO
BACK PORCH CLAMP
5.0 MHZ LPF
AM MODULATOR
38.9 MHZ IF VIDEO CARRIER
+
SAW FILTER
33.15–40.15 MHZ BANDWIDTH
MIXER (UP CONVERTER)
BANDPASS FILTER
MODULATED RF AUDIO / VIDEO
CHANNEL SELECT
Figure 8.21. Typical RF Modulation Implementation for (G) PAL: Zweiton Stereo Audio.
PAL Overview
transmitted in alternate frames, “110” = one 704 kbps data channel. C4 is a “1” if the analog sound is the same as the digital sound. Stereo Audio Encoding The thirty-two 14-bit samples (1 ms of audio, 2’s complement format) per channel are pre-emphasized to the ITU-T J.17 curve. The largest positive or negative sample of the 32 is used to determine which 10 bits of all 32 samples to transmit. Three range bits per channel (R0L, R1L, R2L, and R0R, R1R, R2R) are used to indicate the scaling factor. D13 is the sign bit (“0” = positive). D13–D0 01xxxxxxxxxxxx 001xxxxxxxxxxx 0001xxxxxxxxxx 00001xxxxxxxxx 000001xxxxxxxx 0000001xxxxxxx 0000000xxxxxxx 1111111xxxxxxx 1111110xxxxxxx 111110xxxxxxxx 11110xxxxxxxxx 1110xxxxxxxxxx 110xxxxxxxxxxx 10xxxxxxxxxxxx
FRAME ALIGNMENT WORD
R2–R0 111 110 101 011 101 010 00x 00x 010 100 011 101 110 111
CONTROL BITS
Bits Used D13, D13, D13, D13, D13, D13, D13, D13, D13, D13, D13, D13, D13, D13,
D12–D4 D11–D3 D10–D2 D9–D1 D8–D0 D8–D0 D8–D0 D8–D0 D8–D0 D8–D0 D9–D1 D10–D2 D11–D3 D12–D4
301
A parity bit for the six MSBs of each sample is added, resulting in each sample being 11 bits. The 64 samples are interleaved, generating L0, R0, L1, R1, L2, R2, ... L31, R31, and numbered 0–63. The parity bits are used to convey to the decoder what scaling factor was used for each channel (“signalling-in-parity”). If R2L = “0,” even parity for samples 0, 6, 12, 18, ... 48 is used. If R2L = “1,” odd parity is used. If R2R = “0,” even parity for samples 1, 7, 13, 19, ... 49 is used. If R2R = “1,” odd parity is used. If R1L = “0,” even parity for samples 2, 8, 14, 20, ... 50 is used. If R1L = “1,” odd parity is used. If R1R = “0,” even parity for samples 3, 9, 15, 21, ... 51 is used. If R1R = “1,” odd parity is used. If R0L = “0,” even parity for samples 4, 10, 16, 22, ... 52 is used. If R0L = “1,” odd parity is used. If R0R = “0,” even parity for samples 5, 11, 17, 23, ... 53 is used. If R0R = “1,” odd parity is used.
704 BITS AUDIO DATA
ADDITIONAL DATA BITS
0, 1, 0, 0, 1, 1, 1, 0, C0, C1, C2, C3, C4, AD0, AD1, AD2, AD3, AD4, AD5, AD6, AD7, AD8, AD9, AD10,
A000, A044, A088, ... A001, A045, A089, ... A002, A046, A090, ... A003, A047, A091, ... : A043, A087, A131, ...
Figure 8.22. NICAM 728 Bitstream for One Frame.
A660, A661, A662, A663, A703
302
Chapter 8: NTSC, PAL, and SECAM Overview
The parity of samples 54–63 is normally even. However, they may be modified to transmit two additional bits of information: If CIB0 = “0,” even parity for samples 54, 55, 56, 57, and 58 is used. If CIB0 = “1,” odd parity is used. If CIB1 = “0,” even parity for samples 59, 60, 61, 62, and 63 is used. If CIB1 = “1,” odd parity is used.
The audio data is bit-interleaved as shown in Figure 8.22 to reduce the influence of dropouts. If the bits are numbered 0–703, they are transmitted in the order 0, 44, 88, ... 660, 1, 45, 89, ... 661, 2, 46, 90, ... 703. The whole frame, except the frame alignment word, is exclusive-ORed with a 1-bit pseudo-random binary sequence (PRBS). The PRBS generator is reinitialized after the frame alignment word of each frame so that the first bit of the sequence processes the C0 bit. The polynomial of the PRBS is x9 + x4 + 1 with an initialization word of “111111111.” Actual transmission consists of taking bits in pairs from the 728 kbps bitstream, then generating 356k symbols per second using Differential Quadrature Phase-Shift Keying (DQPSK). If the symbol is “00,” the subcarrier phase is left unchanged. If the symbol is “01,” the subcarrier phase is delayed 90°. If the symbol is “11,” the subcarrier phase is inverted. If the symbol is “10,” the subcarrier phase is advanced 90°. Finally, the signal is spectrum-shaped to a –30 dB bandwidth of ~700 kHz for (I) PAL or ~500 kHz for (B, G) PAL.
Stereo Audio Decoding A PLL locks to the NICAM subcarrier frequency and recovers the phase changes that represent the encoded symbols. The symbols are decoded to generate the 728 kbps bitstream. The frame alignment word is found and the following bits are exclusive-ORed with a locally-generated PRBS to recover the packet. The C0 bit is tested for 8 frames high, 8 frames low behavior to verify it is a NICAM 728 bitstream. The bit-interleaving of the audio data is reversed, and the “signalling-in-parity” decoded: A majority vote is taken on the parity of samples 0, 6, 12, ... 48. If even, R2L = “0”; if odd, R2L = “1.” A majority vote is taken on the parity of samples 1, 7, 13, ... 49. If even, R2R = “0”; if odd, R2R = “1.” A majority vote is taken on the parity of samples 2, 8, 14, ... 50. If even, R1L = “0”; if odd, R1L = “1.” A majority vote is taken on the parity of samples 3, 9, 15, ... 51. If even, R1R = “0”; if odd, R1R = “1.” A majority vote is taken on the parity of samples 4, 10, 16, ... 52. If even, R0L = “0”; if odd, R0L = “1.” A majority vote is taken on the parity of samples 5, 11, 17, ... 53. If even, R0R = “0”; if odd, R0R = “1.”
PAL Overview
303
A majority vote is taken on the parity of samples 54, 55, 56, 57, and 58. If even, CIB0 = “0”; if odd, CIB0 = “1.”
If R0A = “0,” even parity for samples 2, 5, 8, 11, ... 26 is used. If R0A = “1,” odd parity is used.
A majority vote is taken on the parity of samples 59, 60, 61, 62, and 63. If even, CIB1 = “0”; if odd, CIB1 = “1.”
If R0B = “0,” even parity for samples 29, 32, 35, ... 53 is used. If R0B = “1,” odd parity is used.
Any samples whose parity disagreed with the vote are ignored and replaced with an interpolated value. The left channel uses range bits R2L, R1L, and R0L to determine which bits below the sign bit were discarded during encoding. The sign bit is duplicated into those positions to generate a 14-bit sample. The right channel is similarly processed, using range bits R2R, R1R, and R0R. Both channels are then de-emphasized using the J.17 curve.
The audio data is bit-interleaved; however, odd packets contain 64 samples of audio channel 1 while even packets contain 64 samples of audio channel 2. The rest of the processing is the same as for stereo audio.
Dual Mono Audio Encoding Two blocks of thirty-two 14-bit samples (2 ms of audio, 2’s complement format) are preemphasized to the ITU-T J.17 specification. As with the stereo audio, three range bits per block (R0A, R1A, R2A, and R0B, R1B, R2B) are used to indicate the scaling factor. Unlike stereo audio, the samples are not interleaved. If R2A = “0,” even parity for samples 0, 3, 6, 9, ... 24 is used. If R2A = “1,” odd parity is used. If R2B = “0,” even parity for samples 27, 30, 33, ... 51 is used. If R2B = “1,” odd parity is used. If R1A = “0,” even parity for samples 1, 4, 7, 10, ... 25 is used. If R1A = “1,” odd parity is used. If R1B = “0,” even parity for samples 28, 31, 34, ... 52 is used. If R1B = “1,” odd parity is used.
Analog Channel Assignments Tables 8.5 through 8.7 list the channel assignments for VHF, UHF and cable for various PAL systems. Note that cable systems routinely reassign channel numbers to alternate frequencies to minimize interference and provide multiple levels of programming (such as two versions of a premium movie channel: one for subscribers, and one for nonsubscribers during pre-view times).
Luminance Equation Derivation The equation for generating luminance from RGB information is determined by the chromaticities of the three primary colors used by the receiver and what color white actually is. The chromaticities of the RGB primaries and reference white (CIE illuminate D65) are: R:
xr = 0.64
yr = 0.33
zr = 0.03
G:
xg = 0.29
yg = 0.60
zg = 0.11
B:
xb = 0.15
yb = 0.06
zb = 0.79
white: xw = 0.3127 zw = 0.3583
yw = 0.3290
304
Chapter 8: NTSC, PAL, and SECAM Overview
Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
Channel
(B) PAL, Australia, 7 MHz Channel 0 1 2 3 4 5 5A 6 7 8 9 10 11 12
46.25 57.25 64.25 86.25 95.25 102.25 138.25 175.25 182.25 189.25 196.25 209.25 216.25 223.25
51.75 62.75 69.75 91.75 100.75 107.75 143.75 180.75 187.75 194.75 201.75 214.75 221.75
45.75 53.75 61.75 175.25 183.25 191.25 199.25 207.25 215.25
51.75 59.75 67.75 181.25 189.25 197.25 205.25 213.25 221.25
Audio Carrier (MHz)
Channel Range (MHz)
(B) PAL, Italy, 7 MHz Channel 45–52 56–63 63–70 85–92 94–101 101–108 137–144 174–181 181–188 188–195 195–202 208–215 215–222
A B C D E F G H H–1 H–2 – – – –
(I) PAL, Ireland, 8 MHz Channel 1 2 3 4 5 6 7 8 9
Video Carrier (MHz)
53.75 62.25 82.25 175.25 183.75 192.25 201.25 210.25 217.25 224.25 – – – –
59.25 67.75 87.75 180.75 189.25 197.75 206.75 215.75 222.75 229.75 – – – –
52.5–59.5 61–68 81–88 174–181 182.5–189.5 191–198 200–207 209–216 216–223 223–230 – – – –
(B) PAL, New Zealand, 7 MHz Channel 44.5–52.5 52.5–60.5 60.5–68.5 174–182 182–190 190–198 198–206 206–214 214–222
1 2 3 4 5 6 7 8 9
45.25 55.25 62.25 175.25 182.25 189.25 196.25 203.25 210.25
50.75 60.75 67.75 180.75 187.75 194.75 201.75 208.75 215.75
44–51 54–61 61–68 174–181 181–188 188–195 195–202 202–209 209–216
Table 8.5. Analog Broadcast and Cable TV Nominal Frequencies for (B, I) PAL in Various Countries.
PAL Overview
Broadcast Channel
Audio Carrier (MHz)
Video Carrier (MHz)
Channel Range (MHz)
(G, H) PAL
(I) PAL
21 31 41 51 61 71 81 91 101
45.75 53.75 61.75 175.25 183.25 191.25 199.25 207.25 215.25
51.25 59.25 67.25 180.75 188.75 196.75 204.75 212.75 220.75
51.75 59.75 67.75 181.25 189.25 197.25 205.25 213.25 221.25
44.5–52.5 52.5–60.5 60.5–68.5 174–182 182–190 190–198 198–206 206–214 214–222
22 32 42 52 62 72 82 92 102 112 122
48.25 55.25 62.25 175.25 182.25 189.25 196.25 203.25 210.25 217.25 224.25
53.75 60.75 67.75 180.75 187.75 194.75 201.75 208.75 215.75 222.75 229.75
– – – – – – – – – – –
47–54 54–61 61–68 174–181 181–188 188–195 195–202 202–209 209–216 216–223 223–230
21 22 23 24 25 26 27 28 29
471.25 479.25 487.25 495.25 503.25 511.25 519.25 527.25 535.25
476.75 484.75 492.75 500.75 508.75 516.75 524.75 532.75 540.75
477.25 485.25 493.25 501.25 509.25 517.25 525.25 533.25 541.25
470–478 478–486 486–494 494–502 502–510 510–518 518–526 526–534 534–542
30 31 32 33 34 35 36 37 38 39
543.25 551.25 559.25 567.25 575.25 583.25 591.25 599.25 607.25 615.25
548.75 556.75 564.75 572.75 580.75 588.75 596.75 604.75 612.75 620.75
549.25 557.25 565.25 573.25 581.25 589.25 597.25 605.25 613.25 621.25
542–550 550–558 558–566 566–574 574–582 582–590 590–598 598–606 606–614 614–622
Table 8.6a. Analog Broadcast Nominal Frequencies for the 1 United Kingdom, 1Ireland, 1South Africa, 1Hong Kong, and 2Western Europe.
305
306
Chapter 8: NTSC, PAL, and SECAM Overview
Broadcast Channel
Audio Carrier (MHz)
Video Carrier (MHz)
Channel Range (MHz)
(G, H) PAL
(I) PAL
40 41 42 43 44 45 46 47 48 49
623.25 631.25 639.25 647.25 655.25 663.25 671.25 679.25 687.25 695.25
628.75 636.75 644.75 652.75 660.75 668.75 676.75 684.75 692.75 700.75
629.25 637.25 645.25 653.25 661.25 669.25 677.25 685.25 693.25 701.25
622–630 630–638 638–646 646–654 654–662 662–670 670–678 678–686 686–694 694–702
50 51 52 53 54 55 56 57 58 59
703.25 711.25 719.25 727.25 735.25 743.25 751.25 759.25 767.25 775.25
708.75 716.75 724.75 732.75 740.75 748.75 756.75 764.75 772.75 780.75
709.25 717.25 725.25 733.25 741.25 749.25 757.25 765.25 773.25 781.25
702–710 710–718 718–726 726–734 734–742 742–750 750–758 758–766 766–774 774–782
60 61 62 63 64 65 66 67 68 69
783.25 791.25 799.25 807.25 815.25 823.25 831.25 839.25 847.25 855.25
788.75 796.75 804.75 812.75 820.75 828.75 836.75 844.75 852.75 860.75
789.25 797.25 805.25 813.25 821.25 829.25 837.25 845.25 853.25 861.25
782–790 790–798 798–806 806–814 814–822 822–830 830–838 838–846 846–854 854–862
Table 8.6b. Analog Broadcast Nominal Frequencies for the United Kingdom, Ireland, South Africa, Hong Kong, and Western Europe.
PAL Overview
Cable Channel
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
E2 E3 E4 S 01 S 02 S 03 S1 S2 S3
48.25 55.25 62.25 69.25 76.25 83.25 105.25 112.25 119.25
53.75 60.75 67.75 74.75 81.75 88.75 110.75 117.75 124.75
47–54 54–61 61–68 68–75 75–82 82–89 104–111 111–118 118–125
S S S S S S S S S
S4 S5 S6 S7 S8 S9 S 10 – – –
126.25 133.25 140.75 147.75 154.75 161.25 168.25 – – –
131.75 138.75 145.75 152.75 159.75 166.75 173.75 – – –
125–132 132–139 139–146 146–153 153–160 160–167 167–174 – – –
E5 E6 E7 E8 E9 E 10 E 11 E 12 – – – –
175.25 182.25 189.25 196.25 203.25 210.25 217.25 224.25 – – – –
180.75 187.75 194.75 201.75 208.75 215.75 222.75 229.75 – – – –
174–181 181–188 188–195 195–202 202–209 209–216 216–223 223–230 – – – –
Video Carrier (MHz)
Audio Carrier (MHz)
Channel Range (MHz)
11 12 13 14 15 16 17 18 19
231.25 238.25 245.25 252.25 259.25 266.25 273.25 280.25 287.25
236.75 243.75 250.75 257.75 264.75 271.75 278.75 285.75 292.75
230–237 237–244 244–251 251–258 258–265 265–272 272–279 279–286 286–293
S S S S S S S S S S
20 21 22 23 24 25 26 27 28 29
294.25 303.25 311.25 319.25 327.25 335.25 343.25 351.25 359.25 367.25
299.75 308.75 316.75 324.75 332.75 340.75 348.75 356.75 364.75 372.75
293–300 302–310 310–318 318–326 326–334 334–342 342–350 350–358 358–366 366–374
S S S S S S S S S S S S
30 31 32 33 34 35 36 37 38 39 40 41
375.25 383.25 391.25 399.25 407.25 415.25 423.25 431.25 439.25 447.25 455.25 463.25
380.75 388.75 396.75 404.75 412.75 420.75 428.75 436.75 444.75 452.75 460.75 468.75
374–382 382–390 390–398 398–406 406–414 414–422 422–430 430–438 438–446 446–454 454–462 462–470
Cable Channel
307
Table 8.7. Analog Cable TV Nominal Frequencies for the United Kingdom, Ireland, South Africa, Hong Kong, and Western Europe.
308
Chapter 8: NTSC, PAL, and SECAM Overview
where x and y are the specified CIE 1931 chromaticity coordinates; z is calculated by knowing that x + y + z = 1. As with NTSC, substituting the known values gives us the solution for Kr, Kg, and Kb: Kr Kg Kb
0.3127 ⁄ 0.3290 0.64 0.29 0.15 = 0.33 0.60 0.06 1 0.3583 ⁄ 0.3290 0.03 0.11 0.79
letterboxed image with 430 active lines. On PALplus TVs, it is displayed as a 16:9 picture with 574 active lines, with extended vertical resolution. The full video bandwidth is available for luminance detail. Cross color artifacts are reduced by clean encoding.
–1
Wide Screen Signalling Line 23 contains a Widescreen Signalling (WSS) control signal, defined by ITU-R BT.1119 and ETSI EN 300 294, used by PALplus TVs. This signal indicates:
0.674
= 1.177
Y = (Kr yr)R´ + (Kgyg)G´ + (Kbyb)B´ = (0.674)(0.33)R´ + (1.177)(0.60)G´ + (1.190)(0.06)B´
Program Aspect Ratio: Full Format 4:3 Letterbox 14:9 Center Letterbox 14:9 Top Full Format 14:9 Center Letterbox 16:9 Center Letterbox 16:9 Top Full Format 16:9 Anamorphic Letterbox > 16:9 Center
Y = 0.222R´ + 0.706G´ + 0.071B´
Enhanced services: Camera Mode Film Mode
1.190
Y is defined to be
or
However, the standard Y = 0.299R´ + 0.587G´ + 0.114B´ equation is still used. Adjustments are made in the receiver to minimize color errors.
PALplus PALplus (ITU-R BT.1197 and ETSI ETS 300 731) is the result of a cooperative project started in 1990, undertaken by several European broadcasters. By 1995, they wanted to provide an enhanced definition television system (EDTV), compatible with existing receivers. PALplus has been transmitted by a few broadcasters since 1994. A PALplus picture has a 16:9 aspect ratio. On conventional TVs, it is displayed as a 16:9
Subtitles: Teletext Subtitles Present Open Subtitles Present
PALplus is defined as being Letterbox 16:9 center, camera mode or film mode, helper signals present using modulation, and clean encoding used. Teletext subtitles may or may not be present, and open subtitles may be present only in the active picture area. During a PALplus transmission, any active video on lines 23 and 623 is blanked prior to encoding. In addition to WSS data, line 23 includes 48 ±1 cycles of a 300 ±9 mV subcarrier with a –U phase, starting 51 µs ±250 ns after 0H. Line 623 contains a 10 µs ±250 ns white pulse, starting 20 µs ±250 ns after 0H.
PAL Overview
A PALplus TV has the option of deinterlacing a Film Mode signal and displaying it on a 50-Hz progressive-scan display or using field repeating on a 100-Hz interlaced display. Ghost Cancellation An optional ghost cancellation signal on line 318, defined by ITU-R BT.1124 and ETSI ETS 300 732, allows a suitably adapted TV to measure the ghost signal and cancel any ghosting during the active video. A PALplus TV may or may not support this feature. Vertical Filtering All PALplus sources start out as a 16:9 YCbCr anamorphic image, occupying all 576 active scan lines. Any active video on lines 23 and 623 is blanked prior to encoding (since these lines are used for WSS and reference information), resulting in 574 active lines per frame. Lines 24–310 and 336–622 are used for active video. Before transmission, the 574 active scan lines of the 16:9 image are squeezed into 430 scan lines. To avoid aliasing problems, the vertical resolution is reduced by lowpass filtering. For Y, vertical filtering is done using a Quadrature Mirror Filter (QMF) highpass and lowpass pair. Using the QMF process allows the highpass and lowpass information to be resampled, transmitted, and later recombined with minimal loss. The Y QMF lowpass output is resampled into three-quarters of the original height; little information is lost to aliasing. After clean encoding, it is the letterboxed signal that conventional 4:3 TVs display.
309
The Y QMF highpass output contains the rest of the original vertical frequency. It is used to generate the helper signals that are transmitted using the “black” scan lines not used by the letterbox picture. Film Mode A film mode broadcast has both fields of a frame coming from the same image, as is usually the case with a movie scanned on a telecine. In film mode, the maximum vertical resolution per frame is about 287 cycles per active picture height (cph), limited by the 574 active scan lines per frame. The vertical resolution of Y is reduced to 215 cph so it can be transmitted using only 430 active lines. The QMF lowpass and highpass filters split the Y vertical information into DC– 215 cph and 216–287 cph. The Y lowpass information is re-scanned into 430 lines to become the letterbox image. Since the vertical frequency is limited to a maximum of 215 cph, no information is lost. The Y highpass output is decimated so only one in four lines are transmitted. These 144 lines are used to transmit the helper signals. Because of the QMF process, no information is lost to decimation. The 72 lines above and 72 lines below the central 430-line of the letterbox image are used to transmit the 144 lines of the helper signal. This results in a standard 574 active line picture, but with the original image in its correct aspect ratio, centered between the helper signals. The scan lines containing the 300 mV helper signals are modulated using the U sub-
310
Chapter 8: NTSC, PAL, and SECAM Overview
carrier so they look black and are not visible to the viewer. After Fixed ColorPlus processing, the 574 scan lines are PAL encoded and transmitted as a standard interlaced PAL frame. Camera Mode Camera (or video) mode assumes the fields of a frame are independent of each other, as would be the case when a camera scans a scene in motion. Therefore, the image may have changed between fields. Only intra-field processing is done. In camera mode, the maximum vertical resolution per field is about 143 cycles per active picture height (cph), limited by the 287 active scan lines per field. The vertical resolution of Y is reduced to 107 cph so it can be transmitted using only 215 active lines. The QMF lowpass and highpass filter pair split the Y vertical information into DC–107 cph and 108–143 cph. The Y lowpass information is re-scanned into 215 lines to become the letterbox image. Since the vertical frequency is limited to a maximum of 107 cph, no information is lost. The Y highpass output is decimated so only one in four lines is transmitted. These 72 lines are used to transmit the helper signals. Because of the QMF process, no information is lost to decimation. The 36 lines above and 36 lines below the central 215-line of the letterbox image are used to transmit the 72 lines of the helper signal. This results in a 287 active line picture, but with the original image in its correct aspect ratio, centered between the helper signals. The scan lines containing the 300 mV helper signals are modulated using the U subcarrier so they look black and are not visible to the viewer.
After either Fixed or Motion Adaptive ColorPlus processing, the 287 scan lines are PAL encoded and transmitted as a PAL field. Clean Encoding Only the letterboxed portion of the PALplus signal is clean encoded. The helper signals are not actual PAL video. However, they are close enough to video to pass through the transmission path and remain fairly invisible on standard TVs. ColorPlus Processing Fixed ColorPlus Film Mode uses a Fixed ColorPlus technique, making use of the lack of motion between the two fields of the frame. Fixed ColorPlus depends on the subcarrier phase of the composite PAL signal being of opposite phase when 312 lines apart. If these two lines have the same luminance and chrominance information, it can be separated by adding and subtracting the composite signals from each other. Adding cancels the chrominance, leaving luminance. Subtracting cancels the luminance, leaving chrominance. In practice, Y information above 3 MHz (YHF) is intra-frame averaged since it shares the frequency spectrum with the modulated chrominance. For line [n], YHF is calculated as follows: 0 ≤ n ≤ 214 for 430-line letterboxed image YHF(60 + n) = 0.5(YHF(372 + n) + YHF(60 + n)) YHF(372 + n) = YHF(60 + n)
YHF is then added to the low-frequency Y (YLF) information. The same intra-frame averaging process is also used for Cb and Cr. The 430-line letterbox image is then PAL encoded.
SECAM Overview
Thus, Y information above 3 MHz, and CbCr information, is the same on lines [n] and [n+312]. Y information below 3 MHz may be different on lines [n] and [n+312]. The full vertical resolution of 287 cph is reconstructed by the decoder with the aid of the helper signals. Motion Adaptive ColorPlus (MACP) Camera Mode uses either Motion Adaptive ColorPlus or Fixed ColorPlus, depending on the amount of motion between fields. This requires a motion detector in both the encoder and decoder. To detect movement, the CbCr data on lines [n] and [n+312] are compared. If they match, no movement is assumed, and Fixed ColorPlus operation is used. If the CbCr data doesn’t match, movement is assumed, and Motion Adaptive ColorPlus operation is used. During Motion Adaptive ColorPlus operation, the amount of YHF added to YLF is dependent on the difference between CbCr(n) and CbCr(n + 312). For the maximum CbCr difference, no YHF data for lines [n] and [n+312] is transmitted. In addition, the amount of intra-frame averaged CbCr data mixed with the direct CbCr data is dependent on the difference between CbCr(n) and CbCr(n + 312). For the maximum CbCr difference, only direct CbCr data is transmitted separately for lines [n] and [n+312].
SECAM Overview SECAM (Sequentiel Couleur Avec Mémoire or Sequential Color with Memory) was developed in France, with broadcasting starting in 1967, by realizing that, if color could be bandwidth-limited horizontally, why not also vertically? The two pieces of color information (Db and Dr) added to the monochrome signal
311
could be transmitted on alternate lines, avoiding the possibility of crosstalk. The receiver requires memory to store one line so that it is concurrent with the next line, and also requires the addition of a lineswitching identification technique. Like PAL, SECAM is a 625-line, 50-fieldper-second, 2:1 interlaced system. SECAM was adopted by other countries; however, many are changing to PAL due to the abundance of professional and consumer PAL equipment.
Luminance Information The monochrome luminance (Y) signal is derived from R´G´B´: Y = 0.299R´ + 0.587G´ + 0.114B´
As with NTSC and PAL, the luminance signal occupies the entire video bandwidth. SECAM has several variations, depending on the video bandwidth and placement of the audio subcarrier. The video signal has a bandwidth of 5.0 or 6.0 MHz, depending on the specific SECAM standard.
Color Information SECAM transmits Db information during one line and Dr information during the next line; luminance information is transmitted each line. Db and Dr are scaled versions of B´ – Y and R´ – Y: Dr = –1.902(R´ – Y) Db = 1.505(B´ – Y)
Since there is an odd number of lines, any given line contains Db information on one field and Dr information on the next field. The decoder requires a 1-H delay, switched synchronously with the Db and Dr switching, so
312
Chapter 8: NTSC, PAL, and SECAM Overview
that Db and Dr exist simultaneously in order to convert to YCbCr or RGB.
Color Modulation SECAM uses FM modulation to transmit the Db and Dr color difference information, with each component having its own subcarrier. Db and Dr are lowpass filtered to 1.3 MHz and pre-emphasis is applied. The curve for the pre-emphasis is expressed by:
f 1 + j ------ 85 A = -------------------------f 1 + j --------- 255 where ƒ = signal frequency in kHz. After pre-emphasis, Db and Dr frequency modulate their respective subcarriers. The frequency of each subcarrier is defined as: FOB = 272 FH = 4.250000 MHz (± 2 kHz) FOR = 282 FH = 4.406250 MHz (± 2 kHz)
These frequencies represent no color information. Nominal Dr deviation is ±280 kHz and the nominal Db deviation is ±230 kHz. Figure 8.23 illustrates the frequency modulation process of the color difference signals. The choice of frequency shifts reflects the idea of keeping the frequencies representing critical colors away from the upper limit of the spectrum to minimize distortion. After modulation of Db and Dr, subcarrier pre-emphasis is applied, changing the amplitude of the subcarrier as a function of the frequency deviation. The intention is to reduce the visibility of the subcarriers in areas of low luminance and to improve the signal-to-noise
ratio of highly saturated colors. This preemphasis is given as:
1 + j16F G = M ------------------------1 + j1.26F where F = (ƒ/4286) – (4286/ƒ), ƒ = instantaneous subcarrier frequency in kHz, and 2M = 23 ± 2.5% of luminance amplitude. As shown in Figure 8.24, Db and Dr information is transmitted on alternate scan lines. The initial phase of the color subcarrier is also modified as shown in Table 8.8 to further reduce subcarrier visibility. Note that subcarrier phase information in the SECAM system carries no picture information.
Composite Video Generation The subcarrier data is added to the luminance along with appropriate horizontal and vertical sync signals, blanking signals, and burst signals to generate composite video. As with PAL, SECAM requires some means of identifying the line-switching sequence. Modern practice has been to use a FOR/FOB burst after most horizontal syncs to derive the switching synchronization information, as shown in Figure 8.25.
SECAM Standards Figure 8.26 shows the common designations for SECAM systems. The letters refer to the monochrome standard for line and field rates, video bandwidth (5.0 or 6.0 MHz), audio carrier relative frequency, and RF channel bandwidth. The SECAM refers to the technique to add color information to the monochrome signal. Detailed timing parameters may be found in Table 8.9.
SECAM Overview
Y
RED + DR
CYAN – DR GRAY
ODD LINES YELLOW – DB
BLUE + DB GRAY
EVEN LINES
FOB 4.25
FOR 4.40
6
Figure 8.23. SECAM FM Color Modulation.
The initial phase subcarrier undergoes in each line a variation defined by Frame to frame: 0°, 180°, 0°, 180° ... Line to line: 0°, 0°, 180°, 0°, 0°, 180° ... or 0°, 0°, 0°, 180°, 180°, 180° ...
Table 8.8. SECAM Subcarrier Timing.
MHZ
313
314
Chapter 8: NTSC, PAL, and SECAM Overview START OF VSYNC
DR
DB
620
DR
621
DB
308
DB
622
DR
620
DR
308
309
625
1
311
312
313
3
4
5
6
7
624
625
1
311
312
313
315
316
317
318
319
320
3
4
5
6
7
316
317
318
319
23
320
337
DR
24
DR
315
DR
336
DB
2
314
24
DB
ANALOG FIELD 4
DR
DB
23
ANALOG FIELD 3
623
310
2
314
DR
622
DB
624
DR
ANALOG FIELD 2
310
DB
621
623
DB
309
DR
ANALOG FIELD 1
DB
336
DB
337
Figure 8.24. Four-field SECAM Sequence. See Figure 8.5 for equalization and serration pulse details.
BLANK LEVEL
FO SYNC LEVEL HORIZONTAL BLANKING
Figure 8.25. SECAM Chroma Synchronization Signals.
SECAM Overview
Luminance Equation Derivation The equation for generating luminance from RGB information is determined by the chromaticities of the three primary colors used by the receiver and what color white actually is.
The chromaticities of the RGB primaries and reference white (CIE illuminate D65) are: R:
xr = 0.64
yr = 0.33
zr = 0.03
G:
xg = 0.29
yg = 0.60
zg = 0.11
B:
xb = 0.15
yb = 0.06
zb = 0.79
white: xw = 0.3127 zw = 0.3583
FREQUENCY MODULATED SUBCARRIERS FOR = 282 FH FOB = 272 FH LINE SEQUENTIAL DR AND DB SIGNALS
"D, K, K1, L"
315
"B, G"
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ
LINE / FIELD = 625 / 50 FH = 15.625 KHZ FV = 50 HZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 6.0 MHZ AUDIO CARRIER = 6.5 MHZ CHANNEL BANDWIDTH = 8 MHZ
BLANKING SETUP = 0 IRE VIDEO BANDWIDTH = 5.0 MHZ AUDIO CARRIER = 5.5 MHZ CHANNEL BANDWIDTH: B = 7 MHZ G = 8 MHZ
Figure 8.26. Common SECAM Systems.
yw = 0.3290
316
Chapter 8: NTSC, PAL, and SECAM Overview
where x and y are the specified CIE 1931 chromaticity coordinates; z is calculated by knowing that x + y + z = 1. Once again, substituting the known values gives us the solution for Kr, Kg, and Kb: Kr Kg Kb
0.3127 ⁄ 0.3290 0.64 0.29 0.15 = = 0.33 0.60 0.06 1 0.3583 ⁄ 0.3290 0.03 0.11 0.79 0.674
= 1.177 1.190
–1
Y is defined to be Y = (Kr yr)R´ + (Kgyg)G´ + (Kbyb)B´ = (0.674)(0.33)R´ + (1.177)(0.60)G´ + (1.190)(0.06)B´
or Y = 0.222R´ + 0.706G´ + 0.071B´
However, the standard Y = 0.299R´ + 0.587G´ + 0.114B´ equation is still used. Adjustments are made in the receiver to minimize color errors.
317
SECAM Overview
M
N
525
625
625
FIELD FREQUENCY (FIELDS / SECOND)
59.94
50
50
LINE FREQUENCY (HZ)
15,734
15,625
15,625
PEAK WHITE LEVEL (IRE)
100
100
100
SYNC TIP LEVEL (IRE)
–40
–40 (–43)
–43
7.5 ± 2.5
7.5 ± 2.5 (0)
0
SCAN LINES PER FRAME
SETUP (IRE)
B, G
H
133
I
D, K
K1
L
133
115
115
125
PEAK VIDEO LEVEL (IRE)
120
GAMMA OF RECEIVER
2.2
2.8
2.8
2.8
2.8
2.8
2.8
2.8
VIDEO BANDWIDTH (MHZ)
4.2
5.0 (4.2)
5.0
5.0
5.5
6.0
6.0
6.0
LUMINANCE SIGNAL
1
Y = 0.299R´ + 0.685G´ + 0.114B´ (RGB ARE GAMMA–CORRECTED)
Values in parentheses apply to (NC) PAL used in Argentina.
Table 8.9a. Basic Characteristics of Color Video Signals.
318
Chapter 8: NTSC, PAL, and SECAM Overview
M
N
B, D, G, H, I K, K1, L, NC
63.5555
64
64
Line blanking interval (µs)
10.7 ± 0.1
10.88 ± 0.64
11.85 ± 0.15
0H to start of active video (µs)
9.2 ± 0.1
9.6 ± 0.64
10.5
Front porch (µs)
1.5 ± 0.1
1.92 ± 0.64
1.65 ± 0.15
Line synchronizing pulse (µs)
4.7 ± 0.1
4.99 ± 0.77
4.7 ± 0.2
Rise and fall time of line blanking (10%, 90%) (ns)
140 ± 20
300 ± 100
300 ± 100
Rise and fall time of line synchronizing pulses (10%, 90%) (ns)
140 ± 20
≤ 250
250 ± 50
Characteristics Nominal line period (µs)
Notes: 1. 0H is at 50% point of falling edge of horizontal sync. 2. In case of different standards having different specifications and tolerances, the tightest specification and tolerance is listed. 3. Timing is measured between half-amplitude points on appropriate signal edges.
Table 8.9b. Details of Line Synchronization Signals.
SECAM Overview
M
N
B, D, G, H, I K, K1, L, NC
Field period (ms)
16.6833
20
20
Field blanking interval
20 lines
19–25 lines
25 lines
Rise and fall time of field blanking (10%, 90%) (ns)
140 ± 20
≤ 250
300 ± 100
3H
3H
2.5 H
Equalizing pulse width (µs)
2.3 ± 0.1
2.43 ± 0.13
2.35 ± 0.1
Serration pulse width (µs)
4.7 ± 0.1
4.7 ± 0.8
4.7 ± 0.1
Rise and fall time of synchronizing and equalizing pulses (10%, 90%) (ns)
140 ± 20
< 250
250 ± 50
Characteristics
Duration of equalizing and synchronizing sequences
Notes: 1. In case of different standards having different specifications and tolerances, the tightest specification and tolerance is listed. 2. Timing is measured between half-amplitude points on appropriate signal edges.
Table 8.9c. Details of Field Synchronization Signals.
319
320
Chapter 8: NTSC, PAL, and SECAM Overview
ATTENUATION OF COLOR DIFFERENCE SIGNALS
M / NTSC
M / PAL
B, D, G, H, I, N / PAL
B, D, G, K, K1, K / SECAM
U, V, I, Q: < 2 DB AT 1.3 MHZ > 20 DB AT 3.6 MHZ
< 2 DB AT 1.3 MHZ > 20 DB AT 3.6 MHZ
< 3 DB AT 1.3 MHZ > 20 DB AT 4 MHZ (> 20 DB AT 3.6 MHZ)
< 3 DB AT 1.3 MHZ > 30 DB AT 3.5 MHZ
OR < 2 DB AT < 6 DB AT > 6 DB AT START OF BURST AFTER 0H (µS) BURST DURATION (CYCLES) BURST PEAK AMPLITUDE
(BEFORE LOW FREQUENCY PRE–CORRECTION)
Q: 0.4 MHZ 0.5 MHZ 0.6 MHZ
5.3 ± 0.07
5.8 ± 0.1
5.6 ± 0.1
9 ± 1
9 ± 1
10 ± 1 (9 ± 1)
40 ± 1 IRE
42.86 ± 4 IRE
42.86 ± 4 IRE
Note: Values in parentheses apply to (NC) PAL used in Argentina.
Table 8.9d. Basic Characteristics of Color Video Signals.
Video Test Signals Many industry-standard video test signals have been defined to help test the relative quality of encoding, decoding, and the transmission path, and to perform calibration. Note that some video test signals cannot properly be generated by providing RGB data to an encoder; in this case, YCbCr data may be used. If the video standard uses a 7.5-IRE setup, typically only test signals used for visual examination use the 7.5-IRE setup. Test signals designed for measurement purposes typically use a 0-IRE setup, providing the advantage of defining a known blanking level.
Color Bars Overview Color bars are one of the standard video test signals, and there are several variations, depending on the video standard and application. For this reason, this section reviews the
most common color bar formats. Color bars have two major characteristics: amplitude and saturation. The amplitude of a color bar signal is determined by: (R, G, B) a- × 100 ------------------------------------------------amplitude (%) = max max (R, G, B) b
where max(R,G,B)a is the maximum value of R´G´B´ during colored bars and max(R,G,B)b is the maximum value of R´G´B´ during reference white. The saturation of a color bar signal is less than 100% if the minimum value of any one of the R´G´B´ components is not zero. The saturation is determined by: saturation (%) =
min ( R, G, B ) γ 1 – --------------------------------- × 100 max ( R, G, B )
Video Test Signals
where min(R,G,B) and max(R,G,B) are the minimum and maximum values, respectively, of R´G´B´ during colored bars, and γ is the gamma exponent, typically [1/0.45]. NTSC Color Bars In 1953, it was normal practice for the analog R´G´B´ signals to have a 7.5 IRE setup, and the original NTSC equations assumed this form of input to an encoder. Today, digital R´G´B´ or YCbCr signals typically do not include the 7.5 IRE setup, and the 7.5 IRE setup is added within the encoder. The different color bar signals are described by four amplitudes, expressed in percent, separated by oblique strokes. 100% saturation is implied, so saturation is not specified. The first and second numbers are the white and black amplitudes, respectively. The third and fourth numbers are the white and black amplitudes from which the color bars are derived. For example, 100/7.5/75/7.5 color bars would be 75% color bars with 7.5% setup in which the white bar has been set to 100% and the black bar to 7.5%. Since NTSC systems usually have the 7.5% setup, the two common color bars are 75/7.5/75/7.5 and 100/7.5/100/7.5, which are usually shortened to 75% and 100%, respectively. The 75% bars are most commonly used. Television transmitters do not pass information with an amplitude greater than about 120 IRE. Therefore, the 75% color bars are used for transmission testing. The 100% color bars may be used for testing in situations where a direct connection between equipment is possible. The 75/7.5/75/7.5 color bars are a part of the Electronic Industries Association EIA-189-A Encoded Color Bar Standard. Figure 8.27 shows a typical vectorscope display for full-screen 75% NTSC color bars. Figure 8.28 illustrates the video waveform for 75% color bars.
321
Tables 8.10 and 8.11 list the luminance and chrominance levels for the two common color bar formats for NTSC. For reference, the RGB and YCbCr values to generate the standard NTSC color bars are shown in Tables 8.12 and 8.13. RGB is assumed to have a range of 0–255; YCbCr is assumed to have a range of 16–235 for Y and 16–240 for Cb and Cr. It is assumed any 7.5 IRE setup is implemented within the encoder. PAL Color Bars Unlike NTSC, PAL does not support a 7.5 IRE setup; the black and blank levels are the same. The different color bar signals are usually described by four amplitudes, expressed in percent, separated by oblique strokes. The first and second numbers are the maximum and minimum percentages, respectively, of R´G´B´ values for an uncolored bar. The third and fourth numbers are the maximum and minimum percentages, respectively, of R´G´B´ values for a colored bar. Since PAL systems have a 0% setup, the two common color bars are 100/0/75/0 and 100/0/100/0, which are usually shortened to 75% and 100%, respectively. The 75% color bars are used for transmission testing. The 100% color bars may be used for testing in situations where a direct connection between equipment is possible. The 100/0/75/0 color bars also are referred to as EBU (European Broadcast Union) color bars. All of the color bars discussed in this section are also a part of Specification of Television Standards for 625-line System-I Transmissions (1971) published by the Independent Television Authority (ITA) and the British Broadcasting Corporation (BBC), and ITU-R BT.471.
322
Chapter 8: NTSC, PAL, and SECAM Overview
Figure 8.27. Typical Vectorscope Display for 75% NTSC Color Bars.
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
WHITE
Video Test Signals
+ 100 WHITE LEVEL (100 IRE) + 89 + 77
+ 77 + 72
+ 69 100 IRE
+ 56 + 48 3.58 MHZ COLOR BURST (9 CYCLES)
+ 46 + 36
+ 38
+ 28 + 15
+ 12 20 IRE
BLACK LEVEL (7.5 IRE)
+7
7.5 IRE
BLANK LEVEL (0 IRE) –5
20 IRE
– 16 – 16 40 IRE SYNC LEVEL (– 40 IRE)
Figure 8.28. IRE Values for 75% NTSC Color Bars.
Luminance (IRE)
Chrominance Level (IRE)
Minimum Chrominance Excursion (IRE)
Maximum Chrominance Excursion (IRE)
Chrominance Phase (degrees)
white
76.9
0
–
–
–
yellow
69.0
62.1
37.9
100.0
167.1
cyan
56.1
87.7
12.3
100.0
283.5
green
48.2
81.9
7.3
89.2
240.7
magenta
36.2
81.9
–4.8
77.1
60.7
red
28.2
87.7
–15.6
72.1
103.5
blue
15.4
62.1
–15.6
46.4
347.1
black
7.5
0
–
–
–
Table 8.10. 75/7.5/75/7.5 (75%) NTSC Color Bars.
323
Chapter 8: NTSC, PAL, and SECAM Overview
Luminance (IRE)
Chrominance Level (IRE)
Minimum Chrominance Excursion (IRE)
Maximum Chrominance Excursion (IRE)
Chrominance Phase (degrees)
white
100.0
0
–
–
–
yellow
89.5
82.8
48.1
130.8
167.1
cyan
72.3
117.0
13.9
130.8
283.5
green
61.8
109.2
7.2
116.4
240.7
magenta
45.7
109.2
–8.9
100.3
60.7
red
35.2
117.0
–23.3
93.6
103.5
blue
18.0
82.8
–23.3
59.4
347.1
black
7.5
0
–
–
–
0
191
191
0
0
191
191
191
191
0
0
0
0
B´
191
0
191
0
191
0
191
0
135
135
0
0
Black
0
G´
Blue
191
Red
191
Green
R´
Cyan
Yellow
Magenta
Table 8.11. 100/7.5/100/7.5 (100%) NTSC Color Bars.
White
324
gamma-corrected RGB (gamma = 1/0.45)
linear RGB R
135
135
0
0
G
135
135
135
135
0
0
0
0
B
135
0
135
0
135
0
135
0
YCbCr Y
180
162
131
112
84
65
35
16
Cb
128
44
156
72
184
100
212
128
Cr
128
142
44
58
198
212
114
128
Table 8.12. RGB and YCbCr Values for 75% NTSC Color Bars.
255
0
0
255
255
0
0
G´
255
255
255
255
0
0
0
0
B´
255
0
255
0
255
0
255
0
Blue
Cyan
325
Black
255
Red
Yellow
R´
Green
White
Magenta
Video Test Signals
gamma-corrected RGB (gamma = 1/0.45)
linear RGB R
255
255
0
0
255
255
0
0
G
255
255
255
255
0
0
0
0
B
255
0
255
0
255
0
255
0
81
41
16
YCbCr Y
235
210
170
145
106
Cb Cr
128
16
166
54
202
90
240
128
128
146
16
34
222
240
110
128
Table 8.13. RGB and YCbCr Values for 100% NTSC Color Bars.
Figure 8.29 illustrates the video waveform for 75% color bars. Figure 8.30 shows a typical vectorscope display for full-screen 75% PAL color bars. Tables 8.14, 8.15, and 8.16 list the luminance and chrominance levels for the three common color bar formats for PAL. For reference, the RGB and YCbCr values to generate the standard PAL color bars are shown in Tables 8.17, 8.18, and 8.19. RGB is assumed to have a range of 0–255; YCbCr is assumed to have a range of 16–235 for Y and 16–240 for Cb and Cr.
EIA Color Bars (NTSC) The EIA color bars (Figure 8.28 and Table 8.10) are a part of the EIA-189-A standard. The seven bars (gray, yellow, cyan, green, magenta, red and blue) are at 75% amplitude, 100% saturation. The duration of each color bar is 1/7 of the active portion of the scan line. Note that the black bar in Figure 8.28 and Table 8.10 is not part of the standard and is shown for reference only. The color bar test signal allows checking for hue and color saturation accuracy.
326
Chapter 8: NTSC, PAL, and SECAM Overview
Chrominance Phase (degrees)
Peak-to-Peak Chrominance Luminance (volts) U axis (volts)
V axis (volts)
Total (volts)
Line n (135° burst)
Line n + 1 (225° burst)
–
–
white
0.700
0
–
–
yellow
0.465
0.459
0.105
0.470
167
193
cyan
0.368
0.155
0.646
0.664
283.5
76.5
green
0.308
0.304
0.541
0.620
240.5
119.5
magenta
0.217
0.304
0.541
0.620
60.5
299.5
red
0.157
0.155
0.646
0.664
103.5
256.5
blue
0.060
0.459
0.105
0.470
347
13.0
black
0
0
0
0
–
–
Table 8.14. 100/0/75/0 (75%) PAL Color Bars.
Chrominance Phase (degrees)
Peak-to-Peak Chrominance Luminance (volts) U axis (volts)
V axis (volts)
Total (volts)
Line n (135° burst)
Line n + 1 (225° burst)
white
0.700
0
–
–
–
–
yellow
0.620
0.612
0.140
0.627
167
193
cyan
0.491
0.206
0.861
0.885
283.5
76.5
green
0.411
0.405
0.721
0.827
240.5
119.5
magenta
0.289
0.405
0.721
0.827
60.5
299.5
red
0.209
0.206
0.861
0.885
103.5
256.5
blue
0.080
0.612
0.140
0.627
347
13.0
black
0
0
0
0
–
–
Table 8.15. 100/0/100/0 (100%) PAL Color Bars.
Video Test Signals
Chrominance Phase (degrees)
Peak-to-Peak Chrominance Luminance (volts) U axis (volts)
V axis (volts)
Total (volts)
Line n (135° burst)
Line n + 1 (225° burst)
white
0.700
0
–
–
–
–
yellow
0.640
0.459
0.105
0.470
167
193
cyan
0.543
0.155
0.646
0.664
283.5
76.5
green
0.483
0.304
0.541
0.620
240.5
119.5
magenta
0.392
0.304
0.541
0.620
60.5
299.5
red
0.332
0.155
0.646
0.664
103.5
256.5
blue
0.235
0.459
0.105
0.470
347
13.0
black
0
0
0
0
–
–
+ 100
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
WHITE
Table 8.16. 100/0/100/25 (98%) PAL Color Bars.
WHITE LEVEL (100 IRE) + 88 + 75 + 69
+ 66 100 IRE
+ 53 + 44 4.43 MHZ COLOR BURST (10 CYCLES)
+ 43 + 31
+ 32
+ 22 +9
21.43 IRE +6
BLACK / BLANK LEVEL (0 IRE)
0 21.43 IRE – 13 43 IRE
– 25 – 25 SYNC LEVEL (– 43 IRE)
Figure 8.29. IRE Values for 75% PAL Color Bars.
327
328
Chapter 8: NTSC, PAL, and SECAM Overview
Figure 8.30. Typical Vectorscope Display for 75% PAL Color Bars.
Black
Blue
Red
Magenta
Green
Cyan
Yellow
White
Video Test Signals
gamma-corrected RGB (gamma = 1/0.45) R´
255
191
0
0
191
191
0
0
G´
255
191
191
191
0
0
0
0
B´
255
0
191
0
191
0
191
0
R
255
135
0
0
135
135
0
0
G
255
135
135
135
0
0
0
0
B
255
0
135
0
135
0
135
0
Y
235
162
131
112
84
65
35
16
Cb
128
44
156
72
184
100
212
128
Cr
128
142
44
58
198
212
114
128
linear RGB
YCbCr
0
255
255
0
0
255
255
255
255
0
0
0
0
B´
255
0
255
0
255
0
255
0
255
255
0
0
Black
0
G´
Blue
255
Red
255
Green
Yellow
R´
Cyan
White
Magenta
Table 8.17. RGB and YCbCr Values for 75% PAL Color Bars.
gamma-corrected RGB (gamma = 1/0.45)
linear RGB R
255
255
0
0
G
255
255
255
255
0
0
0
0
B
255
0
255
0
255
0
255
0
YCbCr Y
235
210
170
145
106
81
41
16
Cb
128
16
166
54
202
90
240
128
Cr
128
146
16
34
222
240
110
128
Table 8.18. RGB and YCbCr Values for 100% PAL Color Bars.
329
44
G´
255
255
255
B´
255
44
255
Cyan
Black
44
Blue
255
44
44
44
44
44
44
255
44
255
255
5
5
Red
255
Magenta
R´
Green
Yellow
Chapter 8: NTSC, PAL, and SECAM Overview
White
330
gamma-corrected RGB (gamma = 1/0.45) 255
255
255
44
44
255
linear RGB R
255
255
5
5
G
255
255
255
255
5
5
5
5
B
255
5
255
5
255
5
255
5
YCbCr Y
235
216
186
167
139
120
90
16
Cb
128
44
156
72
184
100
212
128
Cr
128
142
44
58
198
212
114
128
Table 8.19. RGB and YCbCr Values for 98% PAL Color Bars.
EBU Color Bars (PAL) The EBU color bars are similar to the EIA color bars, except a 100 IRE white level is used (see Figure 8.29 and Table 8.14). The six colored bars (yellow, cyan, green, magenta, red, and blue) are at 75% amplitude, 100% saturation, while the white bar is at 100% amplitude. The duration of each color bar is 1/7 of the active portion of the scan line. Note that the black bar in Figure 8.29 and Table 8.14 is not part of the standard and is shown for reference only. The color bar test signal allows checking for hue and color saturation accuracy.
SMPTE Bars (NTSC) This split-field test signal is composed of the EIA color bars for the first 2/3 of the field, the Reverse Blue bars for the next 1/12 of the
field, and the PLUGE test signal for the remainder of the field.
Reverse Blue Bars The Reverse Blue bars are composed of the blue, magenta, and cyan colors bars from the EIA/EBU color bars, but are arranged in a different order—blue, black, magenta, black, cyan, black, and white. The duration of each color bar is 1/7 of the active portion of the scan line. Typically, Reverse Blue bars are used with the EIA or EBU color bar signal in a split-field arrangement, with the EIA/EBU color bars comprising the first 3/4 of the field and the Reverse Blue bars comprising the remainder of the field. This split-field arrangement eases adjustment of chrominance and hue on a color monitor.
Video Test Signals
The NTSC PLUGE test signal (shown in Figure 8.31) is composed of a 7.5 IRE (black level) pedestal with a 40 IRE “–I” phase modulation, a 100 IRE white pulse, a 40 IRE “+Q” phase modulation, and 3.5 IRE, 7.5 IRE, and 11.5 IRE pedestals. Typically, PLUGE is used as part of the SMPTE bars. For PAL, each country has its own slightly different PLUGE configuration, with most differences being in the black pedestal level used, and work is being done on a standard test signal. Figure 8.32 illustrates a typical PAL PLUGE test signal. Usually used as a fullscreen test signal, it is composed of a 0 IRE
PLUGE
WHITE
PLUGE (Picture Line-Up Generating Equipment) is a visual black reference, with one area blacker-than-black, one area at black, and one area lighter-than-black. The brightness of the monitor is adjusted so that the black and blacker-than-black areas are indistinguishable from each other and the lighterthan-black area is slightly lighter (the contrast should be at the normal setting). Additional test signals, such as a white pulse and modulated IQ signals, are usually added to facilitate testing and monitor alignment.
–I
+Q
+ 100
3.58 MHZ COLOR BURST (9 CYCLES)
+ 27.5
+ 27.5
+ 20 + 11.5 BLACK LEVEL (7.5 IRE) + 3.5
– 12.5
BLANK LEVEL (0 IRE)
– 12.5
– 20
SYNC LEVEL (– 40 IRE) MICROSECONDS =
9.7
19.1
28.5
38
331
47
49.6
52
54.5
Figure 8.31. PLUGE Test Signal for NTSC. IRE values are indicated.
Chapter 8: NTSC, PAL, and SECAM Overview
Y Bars The Y bars consist of the luminance-only levels of the EIA/EBU color bars; however, the black level (7.5 IRE for NTSC and 0 IRE for PAL) is included and the color burst is still present. The duration of each luminance bar is therefore 1/8 of the active portion of the scan line. Y bars are useful for color monitor adjustment and measuring luminance nonlinearity. Typically, the Y bars signal is used with the EIA or EBU color bar signal in a split-field arrangement, with the EIA/EBU color bars comprising the first 3/4 of the field and the Y bars signal comprising the remainder of the field.
WHITE LEVEL (IRE)
pedestal with PLUGE (–2 IRE, 0 IRE, and 2 IRE pedestals) and a white pulse. The white pulse may have five levels of brightness (0, 25, 50, 75, and 100 IRE), depending on the scan line number, as shown in Figure 8.32. The PLUGE is displayed on scan lines that have non-zero IRE white pulses. ITU-R BT.1221 discusses considerations for various PAL systems.
+ 100
+ 75
4.43 MHZ COLOR BURST (10 CYCLES)
+ 50
+ 25 + 21.43
FULL DISPLAY LINE NUMBERS
332
63 / 375
115 / 427
167 / 479
219 / 531
+2 0
271 / 583
BLANK / BLACK LEVEL (0 IRE)
–2
– 21.43
SYNC LEVEL (– 43 IRE) MICROSECONDS =
22.5 24.8 27.1 29.4
41
52.6
Figure 8.32. PLUGE Test Signal for PAL. IRE values are indicated.
Video Test Signals
333
Red Field
Modulated Ramp
The Red Field signal consists of a 75% amplitude, 100% saturation red chrominance signal. This is useful as the human eye is sensitive to static noise intermixed in a red field. Distortions that cause small errors in picture quality can be examined visually for the effect on the picture. Typically, the Red Field signal is used with the EIA/EBU color bars signal in a split-field arrangement, with the EIA/EBU color bars comprising the first 3/4 of the field, and the Red Field signal comprising the remainder of the field.
The modulated ramp test signal, shown in Figure 8.34, is composed of a luminance ramp from 0 IRE to either 80 or 100 IRE, superimposed with modulated chrominance that has a phase of 0° ±1° relative to the burst. The 80 IRE ramp provides testing of the normal operating range of the system; a 100 IRE ramp may optionally be used to test the entire operating range. The peak-to-peak modulated chrominance is 40 ±0.5 IRE for (M) NTSC and 42.86 ±0.5 IRE for (B, D, G, H, I) PAL. Note a 0 IRE setup is used. The rise and fall times at the start and end of the modulated ramp envelope are 400 ±25 ns (NTSC systems) or approximately 1 µs (PAL systems). This test signal may be used to measure differential gain. The modulated ramp signal is preferred over a 5step or 10-step modulated staircase signal when testing digital systems.
10-Step Staircase This test signal is composed of ten unmodulated luminance steps of 10 IRE each, ranging from 0 IRE to 100 IRE, shown in Figure 8.33. This test signal may be used to measure luminance nonlinearity.
100 WHITE LEVEL (100 IRE) 90 80 70 60
IRE LEVELS
50 40
COLOR BURST
30 20 10
BLANK LEVEL (0 IRE)
SYNC LEVEL MICROSECONDS =
17.5 21.5 25.5 29.5 33.5 37.5 41.5 45.5 49.5 53.5
61.8
Figure 8.33. Ten-Step Staircase Test Signal for NTSC and PAL.
334
Chapter 8: NTSC, PAL, and SECAM Overview
Modulated Staircase
Modulated Pedestal
The 5-step modulated staircase signal (a 10-step version is also used), shown in Figure 8.35, consists of 5 luminance steps, superimposed with modulated chrominance that has a phase of 0° ±1° relative to the burst. The peakto-peak modulated chrominance amplitude is 40 ±0.5 IRE for (M) NTSC and 42.86 ±0.5 IRE for (B, D, G, H, I) PAL. Note that a 0 IRE setup is used. The rise and fall times of each modulation packet envelope are 400 ±25 ns (NTSC systems) or approximately 1 µs (PAL systems). The luminance IRE levels for the 5-step modulated staircase signal are shown in Figure 8.35. This test signal may be used to measure differential gain. The modulated ramp signal is preferred over a 5-step or 10-step modulated staircase signal when testing digital systems.
The modulated pedestal test signal (also called a three-level chrominance bar), shown in Figure 8.36, is composed of a 50 IRE luminance pedestal, superimposed with three amplitudes of modulated chrominance that has a phase relative to the burst of –90° ±1°. The peak-to-peak amplitudes of the modulated chrominance are 20 ±0.5, 40 ±0.5, and 80 ±0.5 IRE for (M) NTSC and 20 ±0.5, 60 ±0.5, and 100 ±0.5 IRE for (B, D, G, H, I) PAL. Note a 0 IRE setup is used. The rise and fall times of each modulation packet envelope are 400 ±25 ns (NTSC systems) or approximately 1 µs (PAL systems). This test signal may be used to measure chrominance-to-luminance intermodulation and chrominance nonlinear gain.
80 IRE
COLOR BURST
BLANK LEVEL (0 IRE)
SYNC LEVEL MICROSECONDS =
14.9 20.2
51.5 56.7 61.8
Figure 8.34. 80 IRE Modulated Ramp Test Signal for NTSC and PAL.
Video Test Signals
90
72
54
COLOR BURST
36
18
0
BLANK LEVEL (0 IRE)
SYNC LEVEL
Figure 8.35. Five-Step Modulated Staircase Test Signal for NTSC and PAL.
± 40 IRE (± 50) ± 20 IRE (± 30)
± 10 IRE (± 10) 50 IRE COLOR BURST
BLANK LEVEL (0 IRE)
SYNC LEVEL MICROSECONDS =
10.0
17.9
29.8
41.7
53.6
61.6
Figure 8.36. Modulated Pedestal Test Signal for NTSC and PAL. PAL IRE values are shown in parentheses.
335
336
Chapter 8: NTSC, PAL, and SECAM Overview
Multiburst
Line Bar
The multiburst test signal for (M) NTSC, shown in Figure 8.37, consists of a white flag with a peak amplitude of 100 ±1 IRE and six frequency packets, each a specific frequency. The packets have a 40 ±1 IRE pedestal with peak-to-peak amplitudes of 60 ±0.5 IRE. Note a 0 IRE setup is used and the starting and ending point of each packet is at zero phase. The ITU multiburst test signal for (B, D, G, H, I) PAL, shown in Figure 8.38, consists of a 4 µs white flag with a peak amplitude of 80 ±1 IRE and six frequency packets, each a specific frequency. The packets have a 50 ±1 IRE pedestal with peak-to-peak amplitudes of 60 ±0.5 IRE. Note the starting and ending points of each packet are at zero phase. The gaps between packets are 0.4–2.0 µs. The ITU multiburst test signal may be present on line 18. The multiburst signals are used to test the frequency response of the system by measuring the peak-to-peak amplitudes of the packets.
Multipulse The (M) NTSC multipulse contains a 2T pulse and 25T and 12.5T pulses with various high-frequency components, as shown in Figure 8.39. The (B, D, G, H, I) PAL multipulse is similar, except 20T and 10T pulses are used, and there is no 7.5 IRE setup. This test signal is typically used to measure the frequency response of the transmission channel. MHZ (CYCLES)
100 IRE
0.5 (4)
The line bar is a single 100 ±0.5 IRE (reference white) pulse of 10 µs (PAL), 18 µs (NTSC), or 25 µs (PAL) that occurs anywhere within the active scan line time (rise and fall times are ≤ 1 µs). Note that the color burst is not present, and a 0 IRE setup is used. This test signal is used to measure line time distortion (line tilt or H tilt). A digital encoder or decoder does not generate line time distortion; the distortion is generated primarily by the analog filters and transmission channel.
1.25 (8)
2 (10)
3 (14)
3.58 (16)
4.2 (18) 70 IRE
COLOR BURST
40 IRE
10 IRE BLANK LEVEL (0 IRE)
SYNC LEVEL
Figure 8.37. Multiburst Test Signal for NTSC.
Video Test Signals
337
MHZ 0.5
1
2
4
4.8
5.8 80 IRE
50 IRE COLOR BURST
20 IRE
BLANK LEVEL (0 IRE)
SYNC LEVEL MICROSECONDS =
12
20
24
30
36
42
48
54
62
Figure 8.38. ITU Multiburst Test Signal for PAL.
2T
1.0
2.0
25T (20T)
12.5T (10T)
3.0 (4.0)
3.58 (4.8)
4.2 (5.8)
MHZ (MHZ)
12.5T 12.5T 12.5T (10T) (10T) (10T)
COLOR BURST
Figure 8.39. Multipulse Test Signal for NTSC and PAL. PAL values are shown in parentheses.
338
Chapter 8: NTSC, PAL, and SECAM Overview
Field Square Wave
The field square wave contains 100 ±0.5 IRE pulses for the entire active line time for Field 1 and blanked scan lines for Field 2. Note that the color burst is not present and a 0 IRE setup is used. This test signal is used to measure field time distortion (field tilt or V tilt). A digital encoder or decoder does not generate field time distortion; the distortion is generated primarily by the analog filters and transmission channel.
2T
100 IRE
Composite Test Signal NTC-7 Version for NTSC The NTC (U. S. Network Transmission Committee) has developed a composite test signal that may be used to test several video parameters, rather than using multiple test signals. The NTC-7 composite test signal for NTSC systems (shown in Figure 8.40) consists of a 100 IRE line bar, a 2T pulse, a 12.5T chrominance pulse, and a 5-step modulated staircase signal.
12.5T 90
72
54 3.58 MHZ COLOR BURST (9 CYCLES)
36 + 20 18
0 BLANK LEVEL (0 IRE)
- 20
SYNC LEVEL (– 40 IRE) MICROSECONDS =
12
30
34
37
42
46 49 52 55 58 61
Figure 8.40. NTC-7 Composite Test Signal for NTSC, With Corresponding IRE Values.
Video Test Signals
The line bar has a peak amplitude of 100 ±0.5 IRE, and 10–90% rise and fall times of 125 ±5 ns with an integrated sine-squared shape. It has a width at the 60 IRE level of 18 µs. The 2T pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 250 ±10 ns. The 12.5T chrominance pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 1562.5 ±50 ns. The 5-step modulated staircase signal consists of 5 luminance steps superimposed with a 40 ±0.5 IRE subcarrier that has a phase of 0° ±1° relative to the burst. The rise and fall times of each modulation packet envelope are 400 ±25 ns. The NTC-7 composite test signal may be present on line 17.
ITU Version for PAL The ITU (BT.628 and BT.473) has developed a composite test signal that may be used to test several video parameters, rather than using multiple test signals. The ITU composite test signal for PAL systems (shown in Figure 8.41) consists of a white flag, a 2T pulse, and a 5-step modulated staircase signal. The white flag has a peak amplitude of 100 ±1 IRE and a width of 10 µs. The 2T pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 200 ±10 ns. The 5-step modulated staircase signal consists of 5 luminance steps (whose IRE values are shown in Figure 8.41) superimposed with a 42.86 ±0.5 IRE subcarrier that has a phase of 60° ±1° relative to the U axis. The rise and fall times of each modulation packet envelope are approximately 1 µs.
2T
100 IRE
100
80
60 4.43 MHZ COLOR BURST (10 CYCLES)
40
20
0
BLANK LEVEL (0 IRE)
SYNC LEVEL (– 43 IRE) MICROSECONDS =
12
22
26
339
30
40 44 48 52 56 60
Figure 8.41. ITU Composite Test Signal for PAL, With Corresponding IRE Values.
340
Chapter 8: NTSC, PAL, and SECAM Overview
The ITU composite test signal may be present on line 330. U.K. Version The United Kingdom allows the use of a slightly different test signal since the 10T pulse is more sensitive to delay errors than the 20T pulse (at the expense of occupying less chrominance bandwidth). Selection of an appropriate pulse width is a trade-off between occupying the PAL chrominance bandwidth as fully as possible and obtaining a pulse with sufficient sensitivity to delay errors. Thus, the national test signal (developed by the British Broadcasting Corporation and the Independent Television Authority) in Figure 8.42 may be present on lines 19 and 332 for (I) PAL systems in the United Kingdom.
2T
100 IRE
The white flag has a peak amplitude of 100 ±1 IRE and a width of 10 µs. The 2T pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 200 ±10 ns. The 10T chrominance pulse has a peak amplitude of 100 ±0.5 IRE. The 5-step modulated staircase signal consists of 5 luminance steps (whose IRE values are shown in Figure 8.42) superimposed with a 21.43 ±0.5 IRE subcarrier that has a phase of 60° ±1° relative to the U axis. The rise and fall times of each modulation packet envelope is approximately 1 µs.
10T 100
80
60 4.43 MHZ COLOR BURST (10 CYCLES)
40
20
0
BLANK LEVEL (0 IRE)
SYNC LEVEL (– 43 IRE) MICROSECONDS =
12
22
26
30
34
40 44 48 52 56 60
Figure 8.42. United Kingdom (I) PAL National Test Signal #1, With Corresponding IRE Values.
Video Test Signals
Combination Test Signal NTC-7 Version for NTSC The NTC (U. S. Network Transmission Committee) has also developed a combination test signal that may be used to test several video parameters, rather than using multiple test signals. The NTC-7 combination test signal for NTSC systems (shown in Figure 8.43) consists of a white flag, a multiburst, and a modulated pedestal signal. The white flag has a peak amplitude of 100 ±1 IRE and a width of 4 µs. The multiburst has a 50 ±1 IRE pedestal with peak-to-peak amplitudes of 50 ±0.5 IRE. The starting point of each frequency packet is at zero phase. The width of the 0.5 MHz packet is 5 µs; the width of the remaining packets is 3 µs.
The 3-step modulated pedestal is composed of a 50 IRE luminance pedestal, superimposed with three amplitudes of modulated chrominance (20 ±0.5, 40 ±0.5, and 80 ±0.5 IRE peak-to-peak) that have a phase of –90° ±1° relative to the burst. The rise and fall times of each modulation packet envelope are 400 ±25 ns. The NTC-7 combination test signal may be present on line 280. ITU Version for PAL The ITU (BT.473) has developed a combination test signal that may be used to test several video parameters, rather than using multiple test signals. The ITU combination test signal for PAL systems (shown in Figure 8.44) consists of a white flag, a 2T pulse, a 20T modulated chrominance pulse, and a 5-step luminance staircase signal.
100 IRE
MHZ 0.5
1
2
3 3.58 4.2
50 IRE COLOR BURST
BLANK LEVEL (0 IRE)
SYNC LEVEL MICROSECONDS =
12
18
341
24 28 32 36 40
46
50
54
60
Figure 8.43. NTC-7 Combination Test Signal for NTSC.
342
Chapter 8: NTSC, PAL, and SECAM Overview
The line bar has a peak amplitude of 100 ±1 IRE and a width of 10 µs. The 2T pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 200 ±10 ns. The 20T chrominance pulse has a peak amplitude of 100 ±0.5 IRE, with a half-amplitude width of 2.0 ±0.06 µs. The 5-step luminance staircase signal consists of 5 luminance steps, at 20, 40, 60, 80 and 100 ±0.5 IRE. The ITU combination test signal may be present on line 17. ITU ITS Version for PAL The ITU (BT.473) has developed a combination ITS (insertion test signal) that may be used to test several PAL video parameters, rather than using multiple test signals. The ITU combination ITS for PAL systems (shown in Figure 8.45) consists of a 3-step modulated pedestal with peak-to-peak amplitudes of 20, 2T
100 IRE
60, and 100 ±1 IRE, and an extended subcarrier packet with a peak-to-peak amplitude of 60 ±1 IRE. The rise and fall times of each subcarrier packet envelope are approximately 1 µs. The phase of each subcarrier packet is 60° ±1° relative to the U axis. The tolerance on the 50 IRE level is ±1 IRE. The ITU composite ITS may be present on line 331. U. K. Version The United Kingdom allows the use of a slightly different test signal, as shown in Figure 8.46. It may be present on lines 20 and 333 for (I) PAL systems in the United Kingdom. The test signal consists of a 50 IRE luminance bar, part of which has a 100 IRE subcarrier superimposed that has a phase of 60° ±1° relative to the U axis, and an extended burst of subcarrier on the second half of the scan line.
20T
100 IRE
80
60 4.43 MHZ COLOR BURST (10 CYCLES)
40
20
BLANK LEVEL (0 IRE)
SYNC LEVEL (– 43 IRE) MICROSECONDS =
12
22
26
32
40
44 48 52 56
62
Figure 8.44. ITU Combination Test Signal for PAL.
Video Test Signals
100 IRE
80 80 IRE 60
4.43 MHZ COLOR BURST (10 CYCLES)
50 IRE
40 20 IRE 20
BLANK LEVEL (0 IRE)
SYNC LEVEL (– 43 IRE) MICROSECONDS =
12 14
18
22
28
34
60
Figure 8.45. ITU Combination ITS Test Signal for PAL.
100 IRE
4.43 MHZ COLOR BURST (10 CYCLES)
50 IRE
21.43 IRE
BLANK LEVEL (0 IRE)
–21.43 IRE
SYNC LEVEL (– 43 IRE) MICROSECONDS =
12 14
28 32 34
62
Figure 8.46. United Kingdom (I) PAL National Test Signal #2.
343
344
Chapter 8: NTSC, PAL, and SECAM Overview
systems, whereas T, 2T, 10T, and 20T pulses are common for PAL video systems. T is the Nyquist interval or
T Pulse Square waves with fast rise times cannot be used for testing video systems, since attenuation and phase shift of out-of-band components cause ringing in the output signal, obscuring the in-band distortions being measured. T, or sin2, pulses are bandwidth-limited, so are used for testing video systems. The 2T pulse is shown in Figure 8.47 and, like the T pulse, is obtained mathematically by squaring a half-cycle of a sine wave. T pulses are specified in terms of half amplitude duration (HAD), which is the pulse width measured at 50% of the pulse amplitude. Pulses with HADs that are multiples of the time interval T are used to test video systems. As seen in Figures 8.39 through 8.44, T, 2T, 12.5T and 25T pulses are common when testing NTSC video
1/2FC
where FC is the cutoff frequency of the video system. For NTSC, FC is 4 MHz, whereas FC for PAL systems is 5 MHz. Therefore, T for NTSC systems is 125 ns and for PAL systems it is 100 ns. For a T pulse with a HAD of 125 ns, a 2T pulse has a HAD of 250 ns, and so on. The frequency spectra for the 2T pulse is shown in Figure 8.47 and is representative of the energy content in a typical character generator waveform. To generate smooth rising and falling edges of most video signals, a T step (generated by integrating a T pulse) is typically used.
2A
A
90% 240 NS A
250 NS
50%
10% TIME (NS) –250
0
TIME (NS)
250
–250
(A)
0
250
(B)
1.0 0.8 0.6 0.4 0.2 FREQUENCY (MHZ) 0
1
2
3
4
5
(C)
Figure 8.47. The T Pulse. (a) 2T pulse. (b) 2T step. (c) Frequency spectra of the 2T pulse.
VBI Data
T steps have 10–90% rise/fall times of 0.964T and a well-defined bandwidth. The 2T step generated from a 2T pulse is shown in Figure 8.47. The 12.5T chrominance pulse, illustrated in Figure 8.48, is a good test signal to measure any chrominance-to-luminance timing error since its energy spectral distribution is bunched in two relatively narrow bands. Using this signal detects differences in the luminance and chrominance phase distortion, but not between other frequency groups.
VBI Data VBI (vertical blanking interval) data may be inserted up to about five scan lines into the active picture region to ensure it won't be
deleted by equipment replacing the VBI, by DSS MPEG which deletes the VBI, or by cable systems inserting their own VBI data. This is common practice by Neilsen and others to ensure their programming and commercial tracking data gets through the distribution systems to the receivers. In most cases, this will be unseen since it is masked by the TV’s overscan.
Timecode Two types of time coding are commonly used, as defined by ANSI/SMPTE 12M and IEC 461: longitudinal timecode (LTC) and vertical interval timecode (VITC). The LTC is recorded on a separate audio track; as a result, the analog VCR must use
0.5
50%
1562.5 NS 1.0 TIME (NS)
–1562.5
0
345
1562.5
(A) 50%
1562.5 NS
0.5
TIME (NS) –1562.5
0
1562.5
TIME (NS)
(C)
–0.5
(B)
Figure 8.48. The 12.5T Chrominance Pulse. (a) Luma component. (b) Chroma component. (c) Addition of (a) and (b).
346
Chapter 8: NTSC, PAL, and SECAM Overview
high-bandwidth amplifiers and audio heads. This is due to the time code frequency increasing as tape speed increases, until the point that the frequency response of the system results in a distorted time code signal that may not be read reliably. At slower tape speeds, the time code frequency decreases, until at very low tape speeds or still pictures, the time code information is no longer recoverable. The VITC is recorded as part of the video signal; as a result, the time code information is always available, regardless of the tape speed. However, the LTC allows the time code signal to be written without writing a video signal; the VITC requires the video signal to be changed if a change in time code information is required. The LTC therefore is useful for synchronizing multiple audio or audio/video sources. Frame Dropping If the field rate is 60/1.001 fields per second, straight counting at 60 fields per second yields an error of about 108 frames for each hour of running time. This may be handled in one of three ways: Nondrop frame: During a continuous recording, each time count increases by 1 frame. In this mode, the drop frame flag will be a “0.” Drop frame: To minimize the timing error, the first two frame numbers (00 and 01) at the start of each minute, except for minutes 00, 10, 20, 30, 40, and 50, are omitted from the count. In this mode, the drop frame flag will be a “1.” Drop frame for (M) PAL: To minimize the timing error, the first four frame numbers (00 to 03) at the start of every second minute (even minute numbers) are omitted from the count, except for minutes 00, 20, and 40. In this mode, the drop frame flag will be a “1.”
Even with drop framing, there is a longterm error of about 2.26 frames per 24 hours. This error accumulation is the reason timecode generators must be periodically reset if they are to maintain any correlation to the correct time-of-day. Typically, this “reset-to-realtime” is referred to as a “jam sync” procedure. Some jam sync implementations reset the timecode to 00:00:00.00 and, therefore, must occur at midnight; others allow a true re-sync to the correct time-of-day. One inherent problem with jam sync correction is the interruption of the timecode. Although this discontinuity may be brief, it may cause timecode readers to “hiccup” due to the interruption. Longitudinal Timecode (LTC) The LTC information is transferred using a separate serial interface, using the same electrical interface as the AES/EBU digital audio interface standard, and is recorded on a separate track. The basic structure of the time data is based on the BCD system. Tables 8.20 and 8.21 list the LTC bit assignments and arrangement. Note that the 24-hour clock system is used. LTC Timing The modulation technique is such that a transition occurs at the beginning of every bit period. “1” is represented by a second transition one-half a bit period from the start of the bit. “0” is represented when there is no transition within the bit period (see Figure 8.49). The signal has a peak-to-peak amplitude of 0.5– 4.5V, with rise and fall times of 40 ±10 µs (10% to 90% amplitude points). Because the entire frame time is used to generate the 80-bit LTC information, the bit rate (in bits per second) is determined by:
VBI Data
Bit(s)
Function
Note
0–3
units of frames
347
Bit(s)
Function
Note
58
flag 5
note 5 note 6
4–7
user group 1
59
flag 6
8–9
tens of frames
60–63
user group 8
10
flag 1
note 1
64
sync bit
fixed”0”
11
flag 2
note 2
65
sync bit
fixed”0”
12–15
user group 2
66
sync bit
fixed”1”
16–19
units of seconds
67
sync bit
fixed”1”
20–23
user group 3
68
sync bit
fixed”1”
24–26
tens of seconds
69
sync bit
fixed”1”
27
flag 3
70
sync bit
fixed”1”
28–31
user group 4
note 3
71
sync bit
fixed”1”
32–35
units of minutes
72
sync bit
fixed”1”
36–39
user group 5
73
sync bit
fixed”1”
40–42
tens of minutes
74
sync bit
fixed”1”
43
flag 4
75
sync bit
fixed”1”
44–47
user group 6
note 4
76
sync bit
fixed”1”
48–51
units of hours
77
sync bit
fixed”1”
52–55
user group 7
78
sync bit
fixed”0”
56–57
tens of hours
79
sync bit
fixed”1”
Notes: 1. Drop frame flag. 525-line and 1125-line systems: “1” if frame numbers are being dropped, “0” if no frame dropping is done. 625-line systems: “0.” 2. Color frame flag. 525-line systems: “1” if even units of frame numbers identify fields 1 and 2 and odd units of field numbers identify fields 3 and 4. 625-line systems: “1” if timecode is locked to the video signal in accordance with 8-field sequence and the video signal has the “preferred subcarrier-to-linesync phase.” 1125-line systems: “0.” 3. 525-line and 1125-line systems: Phase correction. This bit shall be put in a state so that every 80-bit word contains an even number of “0”s. 625-line systems: Binary group flag 0. 4. 525-line and 1125-line systems: Binary group flag 0. 625-line systems: Binary group flag 2. 5. Binary group flag 1. 6. 525-line and 1125-line systems: Binary group flag 2. 625-line systems: Phase correction. This bit shall be put in a state so that every 80-bit word contains an even number of “0”s.
Table 8.20. LTC Bit Assignments.
348
Chapter 8: NTSC, PAL, and SECAM Overview
Frames (count 0–29 for 525-line and 1125-line systems, 0–24 for 625-line systems) units of frames (bits 0–3)
4-bit BCD (count 0–9); bit 0 is LSB
tens of frames (bits 8–9)
2-bit BCD (count 0–2); bit 8 is LSB
Seconds units of seconds (bits 16–19)
4-bit BCD (count 0–9); bit 16 is LSB
tens of seconds (bits 24–26)
3-bit BCD (count 0–5); bit 24 is LSB
Minutes units of minutes (bits 32–35)
4-bit BCD (count 0–9); bit 32 is LSB
tens of minutes (bits 40–42)
3-bit BCD (count 0–5); bit 40 is LSB
Hours units of hours (bits 48–51)
4-bit BCD (count 0–9); bit 48 is LSB
tens of hours (bits 56–57)
2-bit BCD (count 0–2); bit 56 is LSB
Table 8.21. LTC Bit Arrangement.
"0"
"1"
Figure 8.49. LTC Data Bit Transition Format.
VBI Data
FC = 80 FV
where FV is the vertical frame rate in frames per second. The 80 bits of time code information are output serially, with bit 0 being first. The LTC word occupies the entire frame time, and the data must be evenly spaced throughout this time. The start of the LTC word occurs at the beginning of line 5 ±1.5 lines for 525-line systems, at the beginning of line 2 ±1.5 lines for 625-line systems, and at the vertical sync timing reference of the frame ±1 line for 1125line systems. Vertical Inter val Time Code (VITC) The VITC is recorded during the vertical blanking interval of the video signal in both fields. Since it is recorded with the video, it can be read in still mode. However, it cannot be rerecorded (or restriped). Restriping requires dubbing down a generation, deleting and inserting a new time code. For YPbPr and Svideo interfaces, VITC is present on the Y signal. For analog RGB interfaces, VITC is present on all three signals. As with the LTC, the basic structure of the time data is based on the BCD system. Tables 8.22 and 8.23 list the VITC bit assignments and arrangement. Note that the 24-hour clock system is used. VITC Cyclic Redundancy Check Eight bits (82–89) are reserved for the code word for error detection by means of cyclic redundancy checking. The generating polynomial, x8 + 1, applies to all bits from 0 to 81, inclusive. Figure 8.50 illustrates implementing the polynomial using a shift register. During passage of timecode data, the multiplexer is in position 0 and the data is output while the CRC calculation is done simultaneously by the shift register. After all the timecode data has been output, the shift register contains the
349
CRC value, and switching the multiplexer to position 1 enables the CRC value to be output. Repeating the process on decoding, the shift register contains all zeros if no errors exist. VITC Timing The modulation technique is such that each state corresponds to a binary state, and a transition occurs only when there is a change in the data between adjacent bits from a “1” to “0” or “0” to “1.” No transitions occur when adjacent bits contain the same data. This is commonly referred to as “non-return to zero” (NRZ). Synchronization bit pairs are inserted throughout the VITC data to assist the receiver in maintaining the correct frequency lock. The bit rate (FC) is defined to be: FC = 115 FH ± 2%
where FH is the horizontal line frequency. The 90 bits of time code information are output serially, with bit 0 being first. For 625i (576i) systems, lines 19 and 332 (or 21 and 334) are commonly used for the VITC. For 525i (480i) systems, lines 14 and 277 are commonly used. For 1125i (1080i) systems, lines 9 and 571 are commonly used. To protect the VITC against drop-outs, it may also be present two scan lines later, although any two nonconsecutive scan lines per field may be used. Figure 8.51 illustrates the timing of the VITC data on the scan line. The data must be evenly spaced throughout the VITC word. The 10% to 90% rise and fall times of the VITC bit data should be 200 ±50 ns (525-line and 625line systems) or 100 ±25 ns (1125-line systems) before adding it to the video signal to avoid possible distortion of the VITC signal by downstream chrominance circuits. In most circumstances, the analog lowpass filters after the video D/A converters should suffice for the filtering.
350
Chapter 8: NTSC, PAL, and SECAM Overview
Bit(s)
Function
Note
Bit(s)
Function
0
sync bit
fixed “1”
42–45
units of minutes
fixed “0”
46–49
user group 5
Note
1
sync bit
2–5
units of frames
50
sync bit
fixed “1”
6–9
user group 1
51
sync bit
fixed “0”
10
sync bit
fixed “1”
52–54
tens of minutes
11
sync bit
fixed “0”
55
flag 4
12–13
tens of frames
56–59
user group 6
14
flag 1
note 1
60
sync bit
fixed “1”
15
flag 2
note 2
61
sync bit
fixed”0”
16–19
user group 2
62–65
units of hours
20
sync bit
fixed “1”
66–69
user group 7
21
sync bit
fixed “0”
70
sync bit
fixed”1” fixed”0”
note 4
22–25
units of seconds
71
sync bit
26–29
user group 3
72–73
tens of hours
30
sync bit
fixed “1”
74
flag 5
note 5
31
sync bit
fixed “0”
75
flag 6
note 6
32–34
tens of seconds
76–79
user group 8
35
flag 3
36–39
user group 4
note 3
40
sync bit
fixed “1”
41
sync bit
fixed “0”
80
sync bit
fixed”1”
81
sync bit
fixed”0”
82–89
CRC group
Notes: 1. Drop frame flag. 525-line and 1125-line systems: “1” if frame numbers are being dropped, “0” if no frame dropping is done. 625-line systems: “0.” 2. Color frame flag. 525-line systems: “1” if even units of frame numbers identify fields 1 and 2 and odd units of field numbers identify fields 3 and 4. 625-line systems: “1” if timecode is locked to the video signal in accordance with 8-field sequence and the video signal has the “preferred subcarrier-to-line-sync phase.” 1125-line systems: “0.” 3. 525-line systems: Field flag. “0” during fields 1 and 3, “1” during fields 2 and 4. 625-line systems: Binary group flag 0. 1125-line systems: Field flag. “0” during field 1, “1” during field 2. 4. 525-line and 1125-line systems: Binary group flag 0. 625-line systems: Binary group flag 2. 5. Binary group flag 1. 6. 525-line and 1125-line systems: Binary group flag 2. 625-line systems: Field flag. “0” during fields 1, 3, 5, and 7, “1” during fields 2, 4, 6, and 8.
Table 8.22. VITC Bit Assignments.
VBI Data
Frames (count 0–29 for 525-line and 1125-line systems, 0–24 for 625-line systems) units of frames (bits 2–5)
4-bit BCD (count 0–9); bit 2 is LSB
tens of frames (bits 12–13)
2-bit BCD (count 0–2); bit 12 is LSB
Seconds units of seconds (bits 22–25)
4-bit BCD (count 0–9); bit 22 is LSB
tens of seconds (bits 32–34)
3-bit BCD (count 0–5); bit 32 is LSB
Minutes units of minutes (bits 42–45)
4-bit BCD (count 0–9); bit 42 is LSB
tens of minutes (bits 52–54)
3-bit BCD (count 0–5); bit 52 is LSB
Hours units of hours (bits 62–65)
4-bit BCD (count 0–9); bit 62 is LSB
tens of hours (bits 72–73)
2-bit BCD (count 0–2); bit 72 is LSB
Table 8.23. VITC Bit Arrangement.
0
DATA IN
MUX XOR 1 D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
Q
D
Figure 8.50. VITC CRC Generation.
Q
DATA + CRC OUT
351
352
Chapter 8: NTSC, PAL, and SECAM Overview
63.556 µS (115 BITS) 10 µS MIN
50.286 µS (90 BITS) 2.1 µS MIN
(19 BITS)
80 ±10 IRE
525 / 59.94 SYSTEMS HSYNC
HSYNC
64 µS (115 BITS) 11.2 µS MIN
49.655 µS (90 BITS) 1.9 µS MIN
(21 BITS)
78 ±7 IRE
625 / 50 SYSTEMS HSYNC
HSYNC
29.63 µS (115 BITS) 2.7 µS MIN
23.18 µS (90 BITS)
(10.5 BITS)
1.5 µS MIN
78 ±7 IRE
1125 / 59.94 SYSTEMS HSYNC
HSYNC
Figure 8.51. VITC Position and Timing.
VBI Data
User Bits Content
Timecode Referenced to External Clock
BGF2
BGF1
BGF0
user defined
no
0
0
0
8-bit character set1
no
0
0
1
user defined
yes
0
1
0
reser ved
unassigned
0
1
1
zone3
no
1
0
0
page / line2
no
1
0
1
date and time zone3
yes
1
1
0
line2
yes
1
1
1
date and time
page /
Notes: 1. Conforming to ISO/IEC 646 or 2022. 2. Described in SMPTE 262M. 3. Described in SMPTE 309M. See Tables 8.25 through 8.27.
Table 8.24. LTC and VITC Binary Group Flag (BGF) Bit Definitions.
1 3 5 7
2 4 6 8
7–BIT ISO:
B1 B2 B3 B4
B5 B6 B7 0
8–BIT ISO:
A1 A2 A3 A4
A5 A6 A7 A8
USER GROUPS
ONE ISO CHARACTER
Figure 8.52. Use of Binary Groups to Describe ISO Characters Coded With 7 or 8 Bits.
353
354
Chapter 8: NTSC, PAL, and SECAM Overview
User Group 8 Bit 3
Bit 2
MJD Flag
0
User Group 7
Bit 1
Bit 0
Bit 3
Bit 2
Bit 1
Bit 0
time zone offset code 00H–3FH
Notes: 1. MJD flag: “0” = YYMMDD format, “1” = MJD format.
Table 8.25. Date and Time Zone Format Coding.
User Group
Assignment
Value
1
D
0–9
day units
2
D
0–3
day units
3
M
0–9
month units
4
M
0, 1
month units
5
Y
0–9
year units
6
Y
0–9
year units
Description
Table 8.26. YYMMDD Date Format.
User Bits The binary group flag (BGF) bits shown in Table 8.24 specify the content of the 32 user bits. The 32 user bits are organized as eight groups of four bits each. The user bits are intended for storage of data by users. The 32 bits may be assigned in any manner without restriction, if indicated as user-defined by the binary group flags. If an 8-bit character set conforming to ISO/IEC 646 or 2022 is indicated by the binary group flags, the characters are to be inserted as shown in Figure 8.52. Note that some user bits will be decoded before the binary group flags are decoded; therefore, the decoder must store the early user data before any processing is done.
When the user groups are used to transfer time zone and date information, user groups 7 and 8 specify the time zone and the format of the date in the remaining six user groups, as shown in Tables 8.25 and 8.27. The date may be either a six-digit YYMMDD format (Table 8.26) or a six-digit modified Julian date (MJD), as indicated by the MJD flag.
EIA–608 Closed Captioning This section reviews EIA–608 closed captioning for the hearing impaired in the United States. Closed captioning and text are transmitted during the blanked active line-time portion of lines 21 and 284. However, due to video editing they may occasionally reside on any line between 21–25 and 284–289.
VBI Data
Code
Hours
Code
Hours
Code
Hours
00
UTC
16
UTC + 10.00
2C
UTC + 09.30
01
UTC – 01.00
17
UTC + 09.00
2D
UTC + 08.30
02
UTC – 02.00
18
UTC + 08.00
2E
UTC + 07.30
03
UTC – 03.00
19
UTC + 07.00
2F
UTC + 06.30
04
UTC – 04.00
1A
UTC – 06.30
30
TP–1
05
UTC – 05.00
1B
UTC – 07.30
31
TP–0
06
UTC – 06.00
1C
UTC – 08.30
32
UTC + 12.45
07
UTC – 07.00
1D
UTC – 09.30
33
reserved
08
UTC – 08.00
1E
UTC – 10.30
34
reserved
09
UTC – 09.00
1F
UTC – 11.30
35
reserved
0A
UTC – 00.30
20
UTC + 06.00
36
reserved
0B
UTC – 01.30
21
UTC + 05.00
37
reserved user defined
0C
UTC – 02.30
22
UTC + 04.00
38
0D
UTC – 03.30
23
UTC + 03.00
39
unknown
0E
UTC – 04.30
24
UTC + 02.00
3A
UTC + 05.30
0F
UTC – 05.30
25
UTC + 01.00
3B
UTC + 04.30
10
UTC – 10.00
26
reserved
3C
UTC + 03.30
11
UTC – 11.00
27
reserved
3D
UTC + 02.30
12
UTC – 12.00
28
TP–3
3E
UTC + 01.30
13
UTC + 13.00
29
TP–2
3F
UTC + 00.30
14
UTC + 12.00
2A
UTC + 11.30
15
UTC + 11.00
2B
UTC + 10.30
Table 8.27. Time Zone Offset Codes.
355
356
Chapter 8: NTSC, PAL, and SECAM Overview
Extended data services (XDS) also may be transmitted during the blanked active line-time portion of line 284. XDS may indicate the program name, time into the show, time remaining to the end, and so on. Note that due to editing before transmission, it may be possible that the caption information is occasionally moved down a scan line or two. Therefore, caption decoders should monitor more than just lines 21 and 284 for caption information. Waveform The data format for both lines consists of a clock run-in signal, a start bit, and two 7-bit plus parity words of ASCII data (per X3.41967). For YPbPr and S-video interfaces, captioning is present on the Y signal. For analog
10.5 ± 0.25 µS
RGB interfaces, captioning is present on all three signals. Figure 8.53 illustrates the waveform and timing for transmitting the closed captioning and XDS information and conforms to the Television Synchronizing Waveform for Color Transmission in Subpart E, Part 73 of the FCC Rules and Regulations and EIA-608. The clock run-in is a 7-cycle sinusoidal burst that is frequency-locked and phase-locked to the caption data and is used to provide synchronization for the decoder. The nominal data rate is 32× FH. However, decoders should not rely on this timing relationship due to possible horizontal timing variations introduced by video processing circuitry and VCRs. After the clock run-in signal, the blanking level is maintained for a two data bit duration, followed by a “1” start bit.
12.91 µS
7 CYCLES OF 0.5035 MHZ (CLOCK RUN–IN)
50 ±2 IRE
TWO 7–BIT + PARITY ASCII CHARACTERS (DATA)
S T A R T
3.58 MHZ COLOR BURST (9 CYCLES)
D0–D6
P A R I T Y
D0–D6
P A R I T Y
BLANK LEVEL
40 IRE
SYNC LEVEL
10.003 ± 0.25 µS
27.382 µS
33.764 µS
Figure 8.53. 525-Line Lines 21 and 284 Closed Captioning Timing.
240–288 NS RISE / FALL TIMES (2T BAR SHAPING)
VBI Data
The start bit is followed by 16 bits of data, composed of two 7-bit + odd parity ASCII characters. Caption data is transmitted using a non– return-to-zero (NRZ) code; a “1” corresponds to the 50 ± 2 IRE level and a “0” corresponds to the blanking level (0–2 IRE). The negativegoing crossings of the clock are coherent with the data bit transitions. Typical decoders specify the time between the 50% points of sync and clock run-in to be 10.5 ±0.5 µs, with a ±3% tolerance on FH, 50 ±12 IRE for a “1” bit, and –2 to +12 IRE for a “0” bit. Decoders must also handle bit rise/fall times of 240–480 ns. NUL characters (00H) should be sent when no display or control characters are being transmitted. This, in combination with the clock run-in, enables the decoder to determine whether captioning or text transmission is being implemented. If using only line 21, the clock run-in and data do not need to be present on line 284. However, if using only line 284, the clock runin and data should be present on both lines 21 and 284; data for line 21 would consist of NUL characters. At the decoder, as shown in Figure 8.54, the display area of a 525-line 4:3 interlaced display is typically 15 rows high and 34 columns wide. The vertical display area begins on lines 43 and 306 and ends on lines 237 and 500. The horizontal display area begins 13 µs and ends 58 µs, after the leading edge of horizontal sync. In text mode, all rows are used to display text; each row contains a maximum of 32 characters, with at least a one-column wide space on the left and right of the text. The only transparent area is around the outside of the text area. In caption mode, text usually appears only on rows 1–4 or 12–15; the remaining rows are usually transparent. Each row contains a maxi-
357
mum of 32 characters, with at least a one-column wide space on the left and right of the text. Some caption decoders support up to 48 columns per row, and up to 16 rows, allowing some customization for the display of caption data. Basic Ser vices There are two types of display formats: text and captioning. In understanding the operation of the decoder, it is easier to visualize an invisible cursor that marks the position where the next character will be displayed. Note that if you are designing a decoder, you should obtain the latest FCC Rules and Regulations and EIA-608 to ensure correct operation, as this section is only a summary. Text Mode Text mode uses 7–15 rows of the display and is enabled upon receipt of the Resume Text Display or Text Restart code. When text mode has been selected, and the text memory is empty, the cursor starts at the top-most row, character 1 position. Once all the rows of text are displayed, scrolling is enabled. With each carriage return received, the top-most row of text is erased, the text is rolled up one row (over a maximum time of 0.433 seconds), the bottom row is erased, and the cursor is moved to the bottom row, character 1 position. If new text is received while scrolling, it is seen scrolling up from the bottom of the display area. If a carriage return is received while scrolling, the rows are immediately moved up one row to their final position. Once the cursor moves to the character 32 position on any row, any text received before a carriage return, preamble address code, or backspace will be displayed at the character 32 position, replacing any previous character at
358
Chapter 8: NTSC, PAL, and SECAM Overview
that position. The Text Restart command erases all characters on the display and moves the cursor to the top row, character 1 position.
Pop-on captioning may use rows 1–4 or 12– 15, and is initiated by the Resume Caption Loading command. The display memory is essentially double-buffered. While memory buffer 1 is displayed, memory buffer 2 is being loaded with caption data. At the receipt of a End of Caption code, memory buffer 2 is displayed while memory buffer 1 is being loaded with new caption data. Paint-on captioning, enabled by the Resume Direct Captioning command, is similar to Pop-on captioning, but no double-buffering is used; caption data is loaded directly into display memory. Three types of control codes (preamble address codes, midrow codes, and miscellaneous control codes) are used to specify the format, location, and attributes of the charac-
Captioning Mode Captioning has several modes available, including roll-up, pop-on, and paint-on. Roll-up captioning is enabled by receiving one of the miscellaneous control codes to select the number of rows displayed. “Roll-up captions, 2 rows” enables rows 14 and 15; “rollup captions, 3 rows” enables rows 13–15, “rollup captions, 4 rows” enables rows 12–15. Regardless of the number of rows enabled, the cursor remains on row 15. Once row 15 is full, the rows are scrolled up one row (at the rate of one dot per frame), and the cursor is moved back to row 15, character 1.
LINE COUNT 45.02 µS (34 CHARACTERS) LINE 43 56 69 82 95 108 121 134 147 160 173 186 199 212 225
ROW 1 2 3 4
CAPTIONS OR INFOTEXT
5 6 7 8
INFOTEXT ONLY
9 10 11 12 13 14
CAPTIONS OR INFOTEXT
15
237
Figure 8.54. Closed Captioning Display Format.
VBI Data
ters. Each control code consists of two bytes, transmitted together on line 21 or line 284. On line 21, they are normally transmitted twice in succession to help ensure correct reception. They are not transmitted twice on line 284 to minimize bandwidth used for captioning. The first byte is a nondisplay control byte with a range of 10H to 1FH; the second byte is a display control byte in the range of 20H to 7FH. At the beginning of each row, a control code is sent to initialize the row. Caption roll-up and text modes allow either a preamble address code or midrow control code at the start of a row; the other caption modes use a preamble address code to initialize a row. The preamble address codes are illustrated in Figure 8.55 and Table 8.28. The midrow codes are typically used within a row to change the color, italics, underline, and flashing attributes and should occur only between words. Color, italics, and underline are controlled by the preamble address and midrow codes; flash on is controlled by a
miscellaneous control code. An attribute remains in effect until another control code is received or the end of row is reached. Each row starts with a control code to set the color and underline attributes (white nonunderlined is the default if no control code is received before the first character on an empty row). The color attribute can be changed only by the midrow code of another color; the italics attribute does not change the color attribute. However, a color attribute turns off the italics attribute. The flash on command does not alter the status of the color, italics, or underline attributes. However, a color or italics midrow control code turns off the flash. Note that the underline color is the same color as the character being underlined; the underline resides on dot row 11 and covers the entire width of the character column. Table 8.29, Figure 8.56, and Table 8.30 illustrate the midrow and miscellaneous control code operation. For example, if it were the end of a caption, the control code could be End
PREAMBLE CONTROL CODE (TRANSMITTED TWICE)
START BIT
NON-DISPLAY CONTROL CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
CAPTION TEXT UP TO 32 CHARACTERS PER ROW
DISPLAY CONTROL CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
IDENTIFICATION CODE, ROW POSITION, INDENT, AND DISPLAY CONDITION INSTRUCTIONS
359
START BIT
FIRST TEXT CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
SECOND TEXT CHARACTER (7 BITS LSB FIRST)
BEGINNING OF DISPLAYED CAPTION
Figure 8.55. Closed Captioning Preamble Address Code Format.
ODD PARITY BIT
360
Chapter 8: NTSC, PAL, and SECAM Overview
Non-display Control Byte
Display Control Byte Row Position
D6
D5
D4
D3
D2
D1
D0
0
0
1
0
1
0
1
0
0
1
CH
1
0 1
1 0
1
1
1
0
0
0
0 1
1 0
1 0
D6
D5
D4
D3
D2
D1
D0
1
0
1
1
1
2
1
0
3
1
1
4
1
0
5
1
1
6
1
0
1
1
1
0
9
1
1
10
7 A
B
C
D
U
8
1
0
11
1
0
12
1
1
13
1
0
14
1
1
15
Notes: 1. U: “0” = no underline, “1” = underline. 2. CH: “0” = data channel 1, “1” = data channel 2.
A
B
C
D
Attribute
0
0
0
0
white
0
0
0
1
green
0
0
1
0
blue
0
0
1
1
cyan
0
1
0
0
red
0
1
0
1
yellow magenta
0
1
1
0
0
1
1
1
italics
1
0
0
0
indent 0, white
1
0
0
1
indent 4, white
1
0
1
0
indent 8, white
1
0
1
1
indent 12, white
1
1
0
0
indent 16, white
1
1
0
1
indent 20, white
1
1
1
0
indent 24, white
1
1
1
1
indent 28, white
Table 8.28. Closed Captioning Preamble Address Codes. In text mode, the indent codes may be used to perform indentation; in this instance, the row information is ignored.
VBI Data
of Caption (transmitted twice). It could be followed by a preamble address code (transmitted twice) to start another line of captioning. Characters are displayed using a dot matrix format. Each character cell is typically 16 samples wide and 26 samples high (16 × 26), as shown in Figure 8.57. Dot rows 2–19 are usually used for actual character outlines. Dot rows 0, 1, 20, 21, 24, and 25 are usually blanked to provide vertical spacing between characters, and underlining is typically done on dot rows 22 and 23. Dot columns 0, 1, 14 and 15 are blanked to provide horizontal spacing between characters, except on dot rows 22 and 23 when the underline is displayed. This results in 12 × 18 characters stored in character ROM. Table 8.31 shows the basic character set. Some caption decoders support multiple character sizes within the 16 × 26 region, including 13 × 16, 13 × 24, 12 × 20, and 12 × 26.
Non-display Control Byte
Not all combinations generate a sensible result due to the limited display area available. Optional Captioning Features Three sets of optional features are available for advanced captioning decoders. Optional Attributes Additional color choices are available for advanced captioning decoders, as shown in Table 8.32. If a decoder doesn’t support semitransparent colors, the opaque colors may be used instead. If a specific background color isn’t supported by a decoder, it should default to the black background color. However, if the black foreground color is supported in a decoder, all the background colors should be implemented. A background attribute appears as a standard space on the display, and the attribute
Display Control Byte Attribute
D6
0
D5
0
D4
1
D3
CH
D2
0
D1
0
D0
1
D6
0
D5
1
361
D4
0
D3
D2
D1
D0
0
0
0
white
0
0
1
green
0
1
0
blue
0
1
1
cyan
1
0
0
1
0
1
yellow
1
1
0
magenta
1
1
1
italics
U
red
Notes: 1. U: “0” = no underline, “1” = underline. 2. CH: “0” = data channel 1, “1” = data channel 2. 3. Italics is implemented as a two-dot slant to the right over the vertical range of the character. Some decoders implement a one dot slant for every four scan lines. Underline resides on dot rows 22 and 23, and covers the entire column width.
Table 8.29. Closed Captioning Midrow Codes.
362
Chapter 8: NTSC, PAL, and SECAM Overview TEXT
START BIT
MID–ROW CONTROL CODE (TRANSMITTED TWICE)
TEXT CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
TEXT CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
NON-DISPLAY CONTROL CHARACTER (7 BITS LSB FIRST)
START BIT
DISPLAY CONTROL CHARACTER (7 BITS LSB FIRST)
ODD PARITY BIT
ODD PARITY BIT
Figure 8.56. Closed Captioning Midrow Code Format. Miscellaneous control codes may also be transmitted in place of the midrow control code.
Non-display Control Byte
Display Control Byte Command
D6
0
0
D5
0
0
D4
1
1
D3
CH
CH
D2
1
1
D1
0
1
D0
F
1
D6
0
0
D5
1
1
D4
0
0
D3
D2
D1
D0
0
0
0
0
resume caption loading
0
0
0
1
backspace
0
0
1
0
reserved
0
0
1
1
reserved
0
1
0
0
delete to end of row
0
1
0
1
roll-up captions, 2 rows
0
1
1
0
roll-up captions, 3 rows
0
1
1
1
roll-up captions, 4 rows
1
0
0
0
flash on
1
0
0
1
resume direct captioning
1
0
1
0
text restart
1
0
1
1
resume text display erase displayed memory
1
1
0
0
1
1
0
1
carriage return
1
1
1
0
erase nondisplayed memory
1
1
1
1
end of caption (flip memories)
0
0
0
1
tab offset (1 column)
0
0
1
0
tab offset (2 columns)
0
0
1
1
tab offset (3 columns)
Notes: 1. F: “0” = line 21, “1” = line 284. CH: “0” = data channel 1, “1” = data channel 2. 2. “Flash on” blanks associated characters for 0.25 seconds once per second.
Table 8.30. Closed Captioning Miscellaneous Control Codes.
VBI Data
DOT ROW LINE 43
0
LINE 306 LINE 44
2
LINE 307 LINE 45
4
BLANK DOT
LINE 308 LINE 46
6
LINE 309 LINE 47
8
CHARACTER DOT
LINE 310 LINE 48
10
LINE 311 LINE 49
12
LINE 312 LINE 50
14
LINE 313 LINE 51
16
LINE 314 LINE 52
18
LINE 315 LINE 53
20
LINE 316 LINE 54
22
LINE 317 LINE 55
24
UNDERLINE
LINE 318
Figure 8.57. Typical 16×26 Closed Captioning Character Cell Format for Row 1.
363
Chapter 8: NTSC, PAL, and SECAM Overview
D3
1
CH
D2
D2
D1
D0
0
0
0
0
0
0
0
1
°
0
0
1
0
1/2
0
0
1
1
¿
0
1
0
0
™
0
1
0
1
¢
0
1
1
0
£
0
1
1
1
music note
1
0
0
0
à
®
0
1
transparent space
1
0
è
1
0
1
1
â
1
1
0
0
ê
1
1
0
1
î
1
1
1
0
ô
1
1
1
1
û
1111
0 0
1110
1 1
1101
1
Special Characters
D3
1100
1
D4
1011
0
D5
1001
1
D6
1000
0
D0
0111
0
D1
0110
0
D4
0101
0
D5
0100
D6
Display Control Byte
1010
Nondisplay Control Byte
D6 D5 D4 D3
364
(
0
8
@
H
P
X
ú
h
p
x
001
!
)
1
9
A
I
Q
Y
a
i
q
y
010
“
á
2
:
B
J
R
Z
b
j
r
z
011
#
+
3
;
C
K
S
[
c
k
s
100
$
,
4
F
N
V
í
f
n
v
ñ
111
‘
/
7
?
G
O
W
ó
g
o
w
D2 D1 D0 000
Table 8.31. Closed Captioning Basic Character Set.
ç
VBI Data
Non-display Control Byte D6
0
D5
0
D4
1
D3
CH
D2
0
D1
0
Display Control Byte D0
0
D6
0
D5
1
D4
0
Background Attribute
D3
D2
D1
0
0
0
white
0
0
1
green
D0
0
1
0
blue
0
1
1
cyan
1
0
0
1
0
1
yellow magenta
1
1
0
1
1
1
365
T
red
black
0
0
1
CH
1
1
1
0
1
0
1
1
0
1
transparent
D6
D5
D4
D3
D2
D1
D0
D6
D5
D4
D3
D2
D1
D0
Foreground Attribute
0
0
1
CH
1
1
1
0
1
0
1
1
1
0
black
1
black underline
Notes: 1. F: “0” = opaque, “1” = semi-transparent. 2. CH: “0” = data channel 1, “1” = data channel 2. 3. Underline resides on dot rows 22 and 23, and covers the entire column width.
Table 8.32. Closed Captioning Optional Attribute Codes.
remains in effect until the end of the row or until another background attribute is received. The foreground attributes provide an eighth color (black) as a character color. As with midrow codes, a foreground attribute code turns off italics and blinking, and the least significant bit controls underlining. Background and foreground attribute codes have an automatic backspace for backward compatibility with current decoders. Thus, an attribute must be preceded by a standard space character. Standard decoders display the space and ignore the attribute. Extended decoders display the space, and on receiving the attribute, backspace, then display a space that changes the color and opacity. Thus, text formatting remains the same regardless of the type of decoder.
Optional Closed Group Extensions To support custom features and characters not defined by the standards, the EIA/CEG maintains a set of code assignments requested by various caption providers and decoder manufacturers. These code assignments (currently used to select various character sets) are not compatible with caption decoders in the United States and videos using them should not be distributed in the U. S. market. Closed group extensions require two bytes. Table 8.33 lists the currently assigned closed group extensions to support captioning in the Asian languages.
366
Chapter 8: NTSC, PAL, and SECAM Overview
Optional Extended Characters An additional 64 accented characters (eight character sets of eight characters each) may be supported by decoders, permitting the display of other languages such as Spanish, French, Portuguese, German, Danish, Italian, Finnish, and Swedish. If supported, these accented characters are available in all caption and text modes. Each of the extended characters incorporates an automatic backspace for backward compatibility with current decoders. Thus, an extended character must be preceded by the standard ASCII version of the character. Standard decoders display the ASCII character and ignore the accented character. Extended decoders display the ASCII character, and on receiving the accented character, backspace,
Non-display Control Byte D6
0
D5
0
D4
1
D3
CH
D2
1
D1
1
then display the accented character. Thus, text formatting remains the same regardless of the type of decoder. Extended characters require two bytes. The first byte is 12H or 13H for data channel one (1AH or 1BH for data channel two), followed by a value of 20H–3FH. Extended Data Ser vices Line 284 may contain extended data service information, interleaved with the caption and text information, as bandwidth is available. In this case, control codes are not transmitted twice, as they may be for the caption and text services. Information is transmitted as packets and operates as a separate unique data channel. Data for each packet may or may not be contig-
Display Control Byte D0
1
D6
0
D5
1
D4
0
D3
D2
D1
D0
0
1
0
0
0
1
0
1
0
1
1
0
0
1
1
1
1
0
0
0
1
0
0
1
1
0
1
0
Background Attribute standard character set (normal size) standard character set (double size) first private character set second private character set People’s Republic of China character set (GB 2312) Korean Standard character set (KSC 5601-1987) first registered character set
Notes: 1. CH: “0” = data channel 1, “1” = data channel 2.
Table 8.33. Closed Captioning Optional Closed Group Extensions.
VBI Data
uous and may be separated into subpackets that can be inserted anywhere space is available in the line 284 information stream. There are four types of extended data characters: Control: Control characters are used as a mode switch to enable the extended data mode. They are the first character of two and have a value of 01F to 0FH. Type: Type characters follow the control character (thus, they are the second character of two) and identify the packet type. They have a value of 01F to 0FH. Checksum: Checksum characters always follow the “end of packet” control character. Thus, they are the second character of two and have a value of 00F to 7FH. Informational: These characters may be ASCII or non-ASCII data. They are transmitted in pairs up to and including 32 characters. A NUL character (00H) is used to ensure pairs of characters are always sent.
Control Characters Table 8.34 lists the control codes. The current class describes a program currently being transmitted. The future class describes a program to be transmitted later. It contains the same information and formats as the current class. The channel class describes non-program-specific information about the channel. The miscellaneous class describes miscellaneous information. The public class transmits data or messages of a public service nature. The undefined class is used in proprietary systems for whatever that system wishes.
367
Type Characters (Current, Future Class) Program Identification Number (01H) This packet uses four characters to specify the program start time and date relative to Coordinated Universal Time (UTC). The format is shown in Table 8.35. Minutes have a range of 0–59. Hours have a range of 0–23. Dates have a range of 1–31. Months have a range of 1–12. “T” indicates if a program is routinely tape delayed for the Mountain and Pacific time zones. The “D,” “L,” and “Z” bits are ignored by the decoder. Program Length (02H) This packet has 2, 4, or 6 characters and indicates the scheduled length of the program and elapsed time for the program. The format is shown in Table 8.36. Minutes and seconds have a range of 0–59. Hours have a range of 0–63. Program Name (03H) This packet contains 2–32 ASCII characters that specify the title of the program. Program Type (04H) This packet contains 2–32 characters that specify the type of program. Each character is assigned a keyword, as shown in Table 8.37. Program Rating (05H) This packet, commonly referred to regarding the “V-chip”, contains the information shown in Table 8.38 to indicate the program rating. V indicates if violence is present. S indicates if sexual situations are present. L indicates if adult language is present. D indicates if sexually suggestive dialog is present.
368
Chapter 8: NTSC, PAL, and SECAM Overview
Control Code
Function
Class
01H 02H
start continue
current
03H 04H
start continue
future
05H 06H
start continue
channel
07H 08H
start continue
miscellaneous
09H 0AH
start continue
public service
0BH 0CH
start continue
reserved
0DH 0EH
start continue
undefined
0FH
end
all
Table 8.34. EIA-608 Control Codes.
D6
D5
D4
D3
D2
D1
D0
Character
1
m5
m4
m3
m2
m1
m0
minute
1
D
h4
h3
h2
h1
h0
hour
1
L
d4
d3
d2
d1
d0
date
1
Z
T
m3
m2
m1
m0
month
Table 8.35. EIA-608 Program Identification Number Format.
VBI Data
D6
D5
D4
D3
D2
D1
D0
1
m5
m4
m3
m2
m1
m0
Character length, minute
1
h5
h4
h3
h2
h1
h0
length, hour
1
m5
m4
m3
m2
m1
m0
elapsed time, minute
1
h5
h4
h3
h2
h1
h0
elapsed time, hour
1
s5
s4
s3
s2
s1
s0
elapsed time, second
0
0
0
0
0
0
0
null character
Table 8.36. EIA-608 Program Length Format.
Code (hex)
Keyword
Code (hex)
Keyword
Code (hex)
Keyword
20 21 22 23 24 25 26 27 28 29 2A 2B 2C 2D 2E 2F
education entertainment movie news religious sports other action advertisement animated anthology automobile awards baseball basketball bulletin
30 31 32 33 34 35 36 37 38 39 3A 3B 3C 3D 3E 3F
business classical college combat comedy commentary concert consumer contemporary crime dance documentary drama elementary erotica exercise
40 41 42 43 44 45 46 47 48 49 4A 4B 4C 4D 4E 4F
fantasy farm fashion fiction food football foreign fund raiser game/quiz garden golf government health high school history hobby
50 51 52 53 54 55 56 57 58 59 5A 5B 5C 5D 5E 5F
hockey home horror information instruction international interview language legal live local math medical meeting military miniseries
60 61 62 63 64 65 66 67 68 69 6A 6B 6C 6D 6E 6F
music mystery national nature police politics premiere prerecorded product professional public racing reading repair repeat review
70 71 72 73 74 75 76 77 78 79 7A 7B 7C 7D 7E 7F
romance science series service shopping soap opera special suspense talk technical tennis travel variety video weather western
Table 8.37. EIA-608 Program Types.
369
370
Chapter 8: NTSC, PAL, and SECAM Overview
Program Audio Services (06H) This packet contains two characters as shown in Table 8.39 to indicate the program audio services available. Program Caption Services (07H) This packet contains 2–8 characters as shown in Table 8.40 to indicate the program caption services available. L2–L0 are coded as shown in Table 8.39. Copy Generation Management System (08H) This CGMS-A (Copy Generation Management System—Analog) packet contains 2 characters as shown in Table 8.41. In the case where either B3 or B4 is a “0,” there is no Analog Protection System (B1 and B2 are “0”). B0 is the analog source bit. Program Aspect Ratio (09H) This packet contains two or four characters as shown in Table 8.42 to indicate the aspect ratio of the program. S0–S5 specify the first line containing active picture information. The value of S0–S5 is calculated by subtracting 22 from the first line containing active picture information. The valid range for the first line containing active picture information is 22–85. E0–E5 specify the last line containing active picture information. The last line containing active video is calculated by subtracting the value of E0–E5 from 262. The valid range for the last line containing active picture information is 199–262. When this packet contains all zeros for both characters, or the packet is not detected, an aspect ratio of 4:3 is assumed. The Q0 bit specifies whether the video is squeezed (“1”) or normal (“0”). Squeezed
video (anamorphic) is the result of compressing a 16:9 aspect ratio picture into a 4:3 aspect ratio picture without cropping side panels. The aspect ratio is calculated as follows: 320 / (E – S) : 1
Program Description (10H–17H) This packet contains 1–8 packet rows, with each packet row containing 0–32 ASCII characters. A packet row corresponds to a line of text on the display. Each packet is used in numerical sequence, and if a packet contains no ASCII characters, a blank line will be displayed. Type Characters (Channel Class) Network Name (01H) This packet uses 2–32 ASCII characters to specify the network name. Network Call Letters (02H) This packet uses four or six ASCII characters to specify the call letters of the channel. When six characters are used, they reflect the over-the-air channel number (2–69) assigned by the FCC. Single-digit channel numbers are preceded by a zero or a null character. Channel Tape Delay (03H) This packet uses two characters to specify the number of hours and minutes the local station typically delays network programs. The format of this packet is shown in Table 8.43. Minutes have a range of 0–59. Hours have a range of 0–23. This delay applies to all programs on the channel that have the “T” bit set in their Program ID packet (Table 8.35).
VBI Data
D6
D5
D4
D3
D2
D1
D0
Character
1
D / a2
a1
a0
r2
r1
r0
MPAA movie rating
1
V
S
L / a3
g2
g1
g0
TV rating
r2–r0:
Movie Rating 000 not applicable 001 G 010 PG 011 PG-13 100 R 101 NC-17 110 X 111 not rated
g2–g0:
USA TV Rating 000 not rated 001 TV-Y 010 TV-Y7 011 TV-G 100 TV-PG 101 TV-14 110 TV-MA 111 not rated
a3–a0:
g2–g0:
Canadian English TV Rating 000 exempt 001 C 010 C8 + 011 G 100 PG 101 14 + 110 18 + 111 reserved
g2–g0:
Canadian French TV Rating 000 exempt 001 G 010 8 ans + 011 13 ans + 100 16 ans + 101 18 ans + 110 reserved 111 reserved
xxx0 LD01 0011 0111 1011 1111
MPAA movie rating USA TV rating Canadian English TV rating Canadian French TV rating reserved reserved
Table 8.38. EIA-608 Program Rating Format.
L2–L0:
000 001 010 011 100 101 110 111
D6
D5
D4
D3
D2
D1
D0
1
L2
L1
L0
T2
T1
T0
main audio program
1
L2
L1
L0
S2
S1
S0
second audio program (SAP)
T2–T0:
000 001 010 011 100 101 110 111
unknown mono simulated stereo true stereo stereo surround data service other none
unknown english spanish french german italian other none
Character
S2–S0:
371
000 001 010 011 100 101 110 111
Table 8.39. EIA-608 Program Audio Services Format.
unknown mono video descriptions non-program audio special effects data service other none
372
Chapter 8: NTSC, PAL, and SECAM Overview
D6
D5
D4
D3
D2
D1
D0
1
L2
L1
L0
F
C
T
FCT: 000 001 010 011 100 101 110 111
Character service code
line 21, data channel 1 captioning line 21, data channel 1 text line 21, data channel 2 captioning line 21, data channel 2 text line 284, data channel 1 captioning line 284, data channel 1 text line 284, data channel 2 captioning line 284, data channel 2 text
Table 8.40. EIA-608 Program Caption Services Format.
D6
D5
D4
D3
D2
D1
D0
1
0
B4
B3
B2
B1
B0
0
0
0
0
0
0
0
Character CGMS null
B4–B3 CGMS–A Services: 00 01 10 11
copying permitted without restriction condition not to be used one generation copy allowed no copying permitted
B2–B1 Analog Protection Services: (APS) 00 01 10 11
no APS pseudo-sync pulse on; color striping off pseudo-sync pulse on; 2-line color striping on pseudo-sync pulse on; 4-line color striping on
Table 8.41. EIA-608 CGMS–A Format.
VBI Data
D6
D5
D4
D3
D2
D1
D0
Character
1
S5
S4
S3
S2
S1
S0
start
1
E5
E4
E3
E2
E1
E0
end
1
–
–
–
–
–
Q0
other
0
0
0
0
0
0
0
null
Table 8.42. EIA-608 Program Aspect Ratio Format.
D6
D5
D4
D3
D2
D1
D0
Character
1
m5
m4
m3
m2
m1
m0
minute
1
–
h4
h3
h2
h1
h0
hour
Table 8.43. EIA-608 Channel Tape Delay Format.
D6
D5
D4
D3
D2
D1
D0
Character
1
m5
m4
m3
m2
m1
m0
minute
1
D
h4
h3
h2
h1
h0
hour
1
L
d4
d3
d2
d1
d0
date
1
Z
T
m3
m2
m1
m0
month
1
–
–
–
D2
D1
D0
day
1
Y5
Y4
Y3
Y2
Y1
Y0
year
Table 8.44. EIA-608 Time of Day Format.
373
374
Chapter 8: NTSC, PAL, and SECAM Overview
Type Characters (Miscellaneous Class) Time of Day (01H) This packet uses six characters to specify the current time of day, month, and date relative to Coordinated Universal Time (UTC). The format is shown in Table 8.44. Minutes have a range of 0–59. Hours have a range of 0–23. Dates have a range of 1–31. Months have a range of 1–12. Days have a range of 1 (Sunday) to 7 (Saturday). Years have a range of 0–63 (added to 1990). “T” indicates if a program is routinely tape delayed for the Mountain and Pacific time zones. “D” indicates whether daylight savings time currently is being observed. “L” indicates whether the local day is February 28th or 29th when it is March 1st UTC. “Z” indicates whether the seconds should be set to zero (to allow calibration without having to transmit the full 6 bits of seconds data). Impulse Capture ID (02H) This packet carries the program start time and length, and can be used to tell a VCR to record this program. The format is shown in Table 8.45. Start and length minutes have a range of 0–59. Start hours have a range of 0–23; length hours have a range of 0–63. Dates have a range of 1–31. Months have a range of 1–12. “T” indicates if a program is routinely tape delayed for the Mountain and Pacific time zones. The “D,” “L.” and “Z” bits are ignored by the decoder. Supplemental Data Location (03H) This packet uses 2–32 characters to specify other lines where additional VBI data may be found. Table 8.46 shows the format. “F” indicates field one (“0”) or field two (“1”). N may have a value of 7–31, and indicates a specific line number.
Local Time Zone (04H) This packet uses two characters to specify the viewer time zone and whether the locality observes daylight savings time. The format is shown in Table 8.47. Hours have a range of 0–23. This is the nominal time zone offset, in hours, relative to UTC. “D” is a “1” when the area is using daylight savings time. Out-of-Band Channel Number (40H) This packet uses two characters to specify a channel number to which all subsequent outof-band packets refer. This is the CATV channel number to which any following out-of-band packets belong to. The format is shown in Table 8.48. Caption (CC) and Text (T) Channels CC1, CC2, T1 and T2 are on line 21. CC3, CC4, T3 and T4 are on line 284. A fifth channel on line 284 carries the Extended Data Services. T1-T4 are similar to CC1-CC4, but take over all or half of the screen to display scrolling text information. CC1 is usually the main caption channel. CC2 or CC3 is occasionally used for supporting a second language version. Closed Captioning for PAL For (M) PAL, caption data may be present on lines 18 and 281, however, it may occasionally reside on any line between 18–22 and 281– 285 due to editing. For (B, D, G, H, I, N, NC) PAL video tapes, caption data may be present on lines 22 and 335, however, it may occasionally reside on any line between 22–26 and 335–339 due to editing. The data format, amplitudes, and rise and fall times match those used in the United States. The timing, as shown in Figure 8.58, is slightly different due to the 625-line horizontal timing.
VBI Data
D6
D5
D4
D3
D2
D1
D0
Character
1
m5
m4
m3
m2
m1
m0
1
D
h4
h3
h2
h1
h0
start, hour
1
L
d4
d3
d2
d1
d0
start, date
1
Z
T
m3
m2
m1
m0
start, month
1
m5
m4
m3
m2
m1
m0
length, minute
1
h5
h4
h3
h2
h1
h0
length, hour
start, minute
Table 8.45. EIA-608 Impulse Capture ID Format.
D6
D5
D4
D3
D2
D1
D0
1
F
N4
N3
N2
N1
N0
Character location
Table 8.46. EIA-608 Supplemental Data Format.
D6
D5
D4
D3
D2
D1
D0
Character
1
D
h4
h3
h2
h1
h0
hour
0
0
0
0
0
0
0
null
Table 8.47. EIA-608 Local Time Zone Format.
D6
D5
D4
D3
D2
D1
D0
Character
1
c5
c4
c3
c2
c1
c0
channel low
1
c11
c10
c9
c8
c7
c6
channel high
Table 8.48. EIA-608 Out-of-Band Channel Number Format.
375
376
Chapter 8: NTSC, PAL, and SECAM Overview
Widescreen Signalling To facilitate the handling of various aspect ratios of program material received by TVs, a widescreen signalling (WSS) system has been developed. This standard allows a WSSenhanced 16:9 TV to display programs in their correct aspect ratio. 625i Systems 625i (576i) systems are based on ITU-R BT.1119 and ETSI EN 300 294. For YPbPr and S-video interfaces, WSS is present on the Y signal. For analog RGB interfaces, WSS is present on all three signals. The Analog Copy Generation Management System (CGMS-A) is also supported by the WSS signal.
10.5 ± 0.25 µS
Data Timing For (B, D, G, H, I, N, NC) PAL, WSS data is normally on line 23, as shown in Figure 8.59. However, due to video editing, WSS data may reside on any line between 23–27. The clock frequency is 5 MHz (±100 Hz). The signal waveform should be a sine-squared pulse, with a half-amplitude duration of 200 ±10 ns. The signal amplitude is 500 mV ±5%. The NRZ data bits are processed by a biphase code modulator, such that one data period equals 6 elements at 5 MHz. Data Content The WSS consists of a run-in code, a start code, and 14 bits of data, as shown in Table 8.49.
13.0 µS
7 CYCLES OF 0.500 MHZ (CLOCK RUN–IN)
50 ±2 IRE
TWO 7–BIT + PARITY ASCII CHARACTERS (DATA)
S T A R T
4.43 MHZ COLOR BURST (10 CYCLES)
D0–D6
P A R I T Y
D0–D6
P A R I T Y
BLANK LEVEL
43 IRE
SYNC LEVEL
10.00 ± 0.25 µS
27.5 µS
34.0 µS
Figure 8.58. 625-Line Lines 22 and 335 Closed Captioning Timing.
240–288 NS RISE / FALL TIMES (2T BAR SHAPING)
VBI Data
Run-In The run-in consists of 29 elements at 5 MHz of a specific sequence, shown in Table 8.49.
To allow automatic selection of the display mode, a 16:9 receiver should support the following minimum requirements: Case 1: The 4:3 aspect ratio picture should be centered on the display, with black bars on the left and right sides.
Start Code The start code consists of 24 elements at 5 MHz of a specific sequence, shown in Table 8.49.
Case 2: The 14:9 aspect ratio picture should be centered on the display, with black bars on the left and right sides. Alternately, the picture may be displayed using the full display width by using a small (typically 8%) horizontal geometrical error.
Group A Data The group A data consists of 4 data bits that specify the aspect ratio. Each data bit generates 6 elements at 5 MHz. b0 is the LSB. Table 8.50 lists the data bit assignments and usage. The number of active lines listed in Table 8.50 are for the exact aspect ratio (a = 1.33, 1.56, or 1.78). The aspect ratio label indicates a range of possible aspect ratios (a) and number of active lines: 4:3 14:9 16:9 >16:9
a ≤ 1.46 1.46 < a ≤ 1.66 1.66 < a ≤ 1.90 a > 1.90
500 MV ±5%
Case 3: The 16:9 aspect ratio picture should be displayed using the full width of the display. Case 4: The >16:9 aspect ratio picture should be displayed as in Case 3 or use the full height of the display by zooming in.
527–576 463–526 405–462 < 405
COLOR BURST
RUN IN
START CODE
DATA (B0 - B13)
29 5 MHZ ELEMENTS
24 5 MHZ ELEMENTS
84 5 MHZ ELEMENTS
BLANK LEVEL
43 IRE
SYNC LEVEL
11.00 ± 0.25 µS
377
27.4 µS
FIgure 8.59. 625-Line Line 23 WSS Timing.
190–210 NS RISE / FALL TIMES (2T BAR SHAPING)
378
Chapter 8: NTSC, PAL, and SECAM Overview
run-in
29 elements at 5 MHz
1 1111 0001 1100 0111 0001 1100 0111 (1F1C 71C7H)
start code
24 elements at 5 MHz
0001 1110 0011 1100 0001 1111 (1E 3C1FH)
24 elements at 5 MHz “0” = 000 111 “1” = 111 000 24 elements at 5 MHz “0” = 000 111 “1” = 111 000 18 elements at 5 MHz “0” = 000 111 “1” = 111 000 18 elements at 5 MHz “0” = 000 111 “1” = 111 000
group A (aspect ratio)
group B (enhanced services)
group C (subtitles)
group D (reserved)
b0, b1, b2, b3
b4, b5, b6, b7 (b7 = “0” since reserved)
b8, b9, b10
b11, b12, b13
Table 8.49. 625-Line WSS Information.
b3, b2, b1, b0
Aspect Ratio Label
Format
Position On 4:3 Display
Active Lines
Minimum Requirements
1000
4:3
full format
–
576
case 1
0001
14:9
letterbox
center
504
case 2
0010
14:9
letterbox
top
504
case 2
1011
16:9
letterbox
center
430
case 3
0100
16:9
letterbox
top
430
case 3
1101
> 16:9
letterbox
center
–
case 4
1110
14:9
full format
center
576
–
0111
16:9
full format (anamorphic)
–
576
–
Table 8.50. 625-Line WSS Group A (Aspect Ratio) Data Bit Assignments and Usage.
VBI Data
Group B Data The group B data consists of four data bits that specify enhanced services. Each data bit generates six elements at 5 MHz. Data bit b4 is the LSB. Bits b5 and b6 are used for PALplus. b4: mode 0 camera mode 1 film mode b5: color encoding 0 normal PAL 1 Motion Adaptive ColorPlus b6: helper signals 0 not present 1 present
Group C Data The group C data consists of three data bits that specify subtitles. Each data bit generates six elements at 5 MHz. Data bit b8 is the LSB. b8: teletext subtitles 0 no 1 yes b10, b9: open subtitles 00 no 01 inside active picture 10 outside active picture 11 reserved
Group D Data The group D data consists of three data bits that specify surround sound and copy protection. Each data bit generates six elements at 5 MHz. Data bit b11 is the LSB. b11: surround sound 0 no 1 yes
379
b12: copyright 0 no copyright asserted or unknown 1 copyright asserted b13: copy protection 0 copying not restricted 1 copying restricted
525i Systems EIA-J CPR–1204 and IEC 61880 define a widescreen signalling standard for 525i (480i) systems. For YPbPr and S-video interfaces, WSS is present on the Y signal. For analog RGB interfaces, WSS is present on all three signals. Data Timing Lines 20 and 283 are used to transmit the WSS information, as shown in Figure 8.60. However, due to video editing, it may reside on any line between 20–24 and 283–287. The clock frequency is FSC/8 or about 447.443 kHz; FSC is the color subcarrier frequency of 3.579545 MHz. The signal waveform should be a sine-squared pulse, with a halfamplitude duration of 2.235 µs ±50 ns. The signal amplitude is 70 ±10 IRE for a “1,” and 0 ±5 IRE for a “0.” Data Content The WSS consists of 2 bits of start code, 14 bits of data, and 6 bits of CRC, as shown in Table 8.51. The CRC used is X6 + X + 1, all preset to “1.” Start Code The start code consists of a “1” data bit followed by a “0” data bit, as shown in Table 8.51.
380
Chapter 8: NTSC, PAL, and SECAM Overview
70 ±10 IRE
COLOR BURST
START CODE
START CODE
"1"
"0"
DATA (B0 - B19)
BLANK LEVEL
40 IRE
SYNC LEVEL
49.1 ± 0.44 µS
11.20 ± 0.30 µS
FIgure 8.60. 525-Line Lines 20 and 283 WSS Timing.
start code
“1”
start code
“0”
word 0
b0, b1
word 1
b2, b3, b4, b5
word 2
b6, b7, b8, b9, b10, b11, b12, b13
CRC
b14, b15, b16, b17, b18, b19
Table 8.51. 525-Line WSS Data Bit Assignments and Usage.
2235 ±50 NS RISE / FALL TIMES (2T BAR SHAPING)
VBI Data
Word 0 Data Word 0 data consists of 2 data bits: b1, b0: 00 01 10 11
4:3 aspect ratio 16:9 aspect ratio 4:3 aspect ratio reserved
normal anamorphic letterbox
Word 1 Data Word 1 data consists of 4 data bits: b5, b4, b3, b2: 0000 copy control information 1111 default
Copy control information is transmitted in Word 2 data when Word 1 data is “0000.” When copy control information is not to be transferred, Word 1 data must be set to the default value “1111.” Word 2 Data Word 2 data consists of 14 data bits. When Word 1 data is “0000,” Word 2 data consists of copy control information. Word 2 copy control data must be transferred at the rate of two or more frames per two seconds. Bits b6 and b7 specify the copy generation management system in an analog signal (CGMS-A). CGMS-A consists of two bits of digital information:
381
Bits b8 and b9 specify the operation of the the Macrovision copy protection signals added to the analog NTSC video signal: b9, b8: 00 01 10 11
PSP off PSP on, 2-line split burst on PSP on, split burst off PSP on, 4-line split burst on
PSP is the Macrovision pseudo-sync pulse operation that, if on, will be present on the composite, s-video and Y (of YPbPr) analog video outputs. Split burst operation inverts the normal phase of the first half of the color burst signal on certain scan lines on the composite and s-video analog video outputs. This Analog Protection System (APS) information must also usually be conveyed via the line 284 Extended Data Services Copy Generation Management System packet discussed in the closed captioning section. Bit b10 specifies whether the source originated from an analog pre-recorded medium. b10: 0 1
not analog pre-recorded medium analog pre-recorded medium
Bits b11, b12, and b13 are reserved and are “000.”
Teletext b7, b6: 00 01 10 11
copying permitted one copy permitted reserved no copying permitted
This CGMS-A information must also usually be conveyed via the line 284 Extended Data Services Copy Generation Management System packet discussed in the closed captioning section.
Teletext allows the transmission of text, graphics, and data. Data may be transmitted on any line, although the VBI interval is most commonly used. The teletext standards are specified by ETSI EN 300 706, ITU-R BT.653 and EIA–516. For YPbPr and S-video interfaces, teletext is present on the Y signal. For analog RGB interfaces, teletext is present on all three signals.
382
Chapter 8: NTSC, PAL, and SECAM Overview
There are many systems that use the teletext physical layer to transmit proprietary information. The advantage is that teletext has already been approved in many countries for broadcast, so certification for a new transmission technique is not required. The data rate for teletext is much higher than that used for closed captioning, approaching up to 7 Mbps in some cases. Therefore, ghost cancellation is needed to recover the transmitted data reliably. There are seven teletext systems defined, as shown in Table 8.52. System B (also known as “World System Teletext” or “WST”) has become the defacto standard and most widely adopted solution. EIA–516, also referred to as NABTS (North American Broadcast Teletext Specification), was used a little in the United States, and was an expansion of the BT.653 525-line system C standard. Figure 8.61 illustrates the teletext data on a scan line. If a line normally contains a color burst signal, it will still be present if teletext data is present. The 16 bits of clock run-in (or
Parameter
System A
clock sync) consists of alternating “1’s” and “0’s.” Figures 8.62 and 8.63 illustrates the structure of teletext systems B and C, respectively. System B Teletext Over view Since teletext System B is the defacto teletext standard, a basic overview is presented here. A teletext service typically consists of pages, with each page corresponding to a screen of information. The pages are transmitted one at time, and after all pages have been transmitted, the cycle repeats, with a typical cycle time of about 30 seconds. However, the broadcaster may transmit some pages more frequently than others, if desired. The teletext service is usually based on up to eight magazines (allowing up to eight independent teletext services), with each magazine containing up to 100 pages. Magazine 1 uses page numbers 100–199, magazine 2 uses page numbers 200–299, etc. Each page may also have sub-pages, used to extend the number of pages within a magazine.
System B
System C
System D
625-Line Video Systems bit rate (Mbps)
6.203125
6.9375
5.734375
5.6427875
data amplitude
67 IRE
66 IRE
70 IRE
70 IRE
40 bytes
45 bytes
36 bytes
37 bytes
data per line
525-Line Video Systems bit rate (Mbps)
–
5.727272
5.727272
5.727272
data amplitude
–
70 IRE
70 IRE
70 IRE
data per line
–
37 bytes
36 bytes
37 bytes
Table 8.52. Summary of Teletext Systems and Parameters.
VBI Data
Each page contains 24 rows, with up to 40 characters per row. A character may be a letter, number, symbol, or simple graphic. There are also control codes to select colors and other attributes such as blinking and double height. In addition to teletext information, the teletext protocol may be used to transmit other information, such as subtitling, program delivery control (PDC), and private data. Subtitling Subtitling is similar to the closed captioning used in the United States. “Open” subtitles are the insertion of text directly into the picture prior to transmission. “Closed” subtitles are transmitted separately from the picture. The transmission of closed subtitles in the UK uses teletext page 888. In the case where multiple languages are transmitted using teletext, separate pages are used for each language. Program Delivery Control (PDC) Program Delivery Control (defined by ETSI EN 300 231 and ITU-R BT.809) is a system that controls VCR recording using teletext
CLOCK RUN-IN
383
information. The VCR can be programmed to look for and record various types of programs or a specific program. Programs are recorded even if the transmission time changes for any reason. There are two methods of transmitting PDC information via teletext: methods A and B. Method A places the data on a viewable teletext page, and is usually transmitted on scan line 16. This method is also known as the Video Programming System (VPS). Method B places the data on a hidden packet (packet 26) in the teletext signal. This packet 26 data contains the data on each program, including channel, program data, and start time. Data Broadcasting Data broadcasting may be used to transmit information to private receivers. Typical applications include real-time financial information, airport flight schedules for hotels and travel agents, passenger information for railroads, software upgrades, etc.
DATA AND ADDRESS
COLOR BURST
BLANK LEVEL
SYNC LEVEL
FIgure 8.61. Teletext Line Format.
384
Chapter 8: NTSC, PAL, and SECAM Overview
TELETEXT
APPLICATION LAYER PRESENTATION LAYER
NEXT HEADER PACKET LAST PACKET OF PAGE PACKET 26 PACKET 28
SESSION LAYER
PACKET 27
PAGE
HEADER PACKET
NEXT HEADER PACKET PAGE ADDRESS (1 BYTE)
TRANSPORT LAYER PACKET 27 HEADER PACKET
DATA GROUP
MAGAZINE / PACKET ADDRESS (2 BYTES) DATA BLOCK
NETWORK LAYER
BYTE SYNC (1 BYTE) DATA PACKET
LINK LAYER
CLOCK SYNC (2 BYTES) DATA UNIT
PHYSICAL LAYER
FIgure 8.62. Teletext System B Structure.
VBI Data
TELETEXT
APPLICATION LAYER
ITU-T T.101, ANNEX D
PRESENTATION LAYER
RECORD N RECORD HEADER
SESSION LAYER RECORD 1
DATA GROUP N DATA GROUP HEADER (8 BYTES)
TRANSPORT LAYER DATA GROUP 1
P HEADER (5 BYTES)
S (1 BYTE)
DATA BLOCK
NETWORK LAYER
BYTE SYNC (1 BYTE) DATA PACKET
LINK LAYER
CLOCK SYNC (2 BYTES) DATA UNIT
FIgure 8.63. Teletext System C Structure.
PHYSICAL LAYER
385
386
Chapter 8: NTSC, PAL, and SECAM Overview
Packets 0–23 A typical teletext page uses 24 packets, numbered 0–23, that correspond to the 24 rows on a displayed page. Packet 24 can add a status row at the bottom for user prompting. For each packet, three bits specify the magazine address (1–8), and five bits specify the row address (0–23). The magazine and row address bits are Hamming error protected to permit single-bit errors to be corrected. To save bandwidth, the whole address isn’t sent with all packets. Only packet 0 (also called the header packet) has all the address information such as row, page, and magazine address data. Packets 1–28 contain information that is part of the page identified by the most recent packet 0 of the same magazine. The transmission of a page starts with a header packet. Subsequent packets with the same magazine address provide additional data for that page. These packets may be transmitted in any order, and interleaved with packets from other magazines. A page is considered complete when the next header packet for that magazine is received. The general format for packet 0 is: clock run-in framing code magazine and row address page number subcode control codes display data
2 bytes 1 byte 2 bytes 2 bytes 4 bytes 2 bytes 32 bytes
The general format for packets 1–23 is: clock run-in framing code magazine and row address display data
2 bytes 1 byte 2 bytes 40 bytes
Packet 24 This packet defines an additional row for user prompting. Teletext decoders may use the data in packet 27 to react to prompts in the packet 24 display row. Packet 25 This packet defines a replacement header line. If present, the 40 bytes of data are displayed instead of the channel, page, time, and date from packet 8.30. Packet 26 Packet 26 consists of: clock run-in 2 bytes framing code 1 byte magazine and row address 2 bytes designation code 1 byte 13 3-byte data groups, each consisting of 7 data bits 6 address bits 5 mode bits 6 Hamming bits
There are 15 variations of packet 26, defined by the designation code. Each of the 13 data groups specify a specific display location and data relating to that location. This packet is also used to extend the addressable range of the basic character set in order to support other languages, such as Arabic, Spanish, Hungarian, Chinese, etc. For PDC, packet 26 contains data for each program, identifying the channel, program date, start time, and the cursor position of the program information on the page. When the user selects a program, the cursor position is linked to the appropriate packet 26 preselection data. This data is then used to program the VCR. When the program is transmitted, the
VBI Data
387
program information is transmitted using packet 8.30 format 2. A match between the preselection data and the packet 8.30 data turns the VCR record mode on.
The sixth link corresponds to the Fastext index key on the remote, and specifies the page address to go to when the index is selected.
Packet 27 Packet 27 tells the teletext decoder how to respond to user selections for packet 24. There may be up to four packet 27s (packets 27/0 through 27/3), allowing up to 24 links. Packet 27 consists of:
Packets 28 and 29 These are used to define level 2 and level 3 pages to support higher resolution graphics, additional colors, alternate character sets, etc. They are similar in structure to packet 26.
clock run-in framing code magazine and row address designation code link 1 (red) link 2 (green) link 3 (yellow) link 4 (cyan) link 5 (next page) link 6 (index) link control data page check digit
2 bytes 1 byte 2 bytes 1 byte 6 bytes 6 bytes 6 bytes 6 bytes 6 bytes 6 bytes 1 byte 2 bytes
Each link consists of: 7 data bits 6 address bits 5 mode bits 6 hamming bits
This packet contains information linking the current page to six page numbers (links). The four colored links correspond to the four colored Fastext page request keys on the remote. Typically, these four keys correspond to four colored menu selections at the bottom of the display using packet 24. Selection of one of the colored page request keys results in the selection of the corresponding linked page. The fifth link is used for specifying a page the user might want to see after the current page, such as the next page in a sequence.
Packet 8.30 Format 1 Packet 8.30 (magazine 8, packet 30) isn’t associated with any page, but is sent once per second. This packet is also known as the Television Service Data Packet, or TSDP. It contains data that notifies the teletext decoder about the transmission in general and the time. clock run-in framing code magazine and row address designation code initial teletext page network ID time offset from UTC date (Modified Julian Day) UTC time TV program label status display
2 bytes 1 byte 2 bytes 1 byte 6 bytes 2 bytes 1 byte 3 bytes 3 bytes 4 bytes 20 bytes
The Designation Code indicates whether the transmission is during the VBI or full-field. Initial Teletext Page tells the decoder which page should be captured and stored on power-up. This is usually an index or menu page. The Network Identification code identifies the transmitting network. The TV Program Label indicates the program label for the current program. Status Display is used to display a transmission status message.
388
Chapter 8: NTSC, PAL, and SECAM Overview
Packet 8.30 Format 2 This format is used for PDC recorder control, and is transmitted once per second per stream. It contains a program label indicating the start of each program, usually transmitted about 30 seconds before the start of the program to allow the VCR to detect it and get ready to record. clock run-in framing code magazine and row address designation code initial teletext page label channel ID program control status country and network ID program ID label country and network ID program type status display
2 bytes 1 byte 2 bytes 1 byte 6 bytes 1 byte 1 byte 2 bytes 5 bytes 2 bytes 2 bytes 20 bytes
The content is the same as for Format 1, except for the 13 bytes of information before the status display information. Label channel ID (LCI) identifies each of up to four PDC streams that may be transmitted simultaneously. The Program Control Status (PCS) indicates real-time status information, such as the type of analog sound transmission. The Country and Network ID (CNI) is split into two groups. The first part specifies the country and the second part specifies the network. Program ID Label (PIL) specifies the month, day, and local time of the start of the program.
Program Type (PTY) is a code that indicates an intended audience or a particular series. Examples are “adult,” “children,” “music,” “drama,” etc. Packet 31 Packet 31 is used for the transmission of data to private receivers. It consists of: clock run-in framing code data channel group message bits format type address length address repeat indicator continuity indicator data length user data CRC
2 bytes 1 byte 1 byte 1 byte 1 byte 1 byte 0–6 bytes 0–1 byte 0–1 byte 0–1 byte 28–36 bytes 2 bytes
“Raw” VBI Data “Raw,” or oversampled, VBI data is simply digitized VBI data. It is typically oversampled using a 2× video sample clock, such as 27 MHz for 480i and 54 MHz for 480p video. Use of the 2× video sample clock enables transferring the “raw” VBI data over a standard 8-bit BT.656 interface. VBI data may be present on any scan line, except during the serration and equalization intervals. The “raw” VBI data is then converted to binary (or “sliced”) data and processed and/or passed through to the composite, s-video and YPbPr analog video outputs so it may be decoded by the TV.
VBI Data
In the conversion from “raw” to “sliced” VBI data, the VBI decoders must compensate for varying DC offsets, amplitude variations, ghosting, and timing variations. Hysteresis must also be used to prevent the VBI decoders from turning on and off rapidly due to noise and transmission errors. Once the desired VBI signal is found for (typically) 32 consecutive frames, VBI decoding should commence. When the desired VBI signal is not found on the appropriate scan lines for (typically) 32 consecutive frames, VBI decoding should stop.
“Sliced” VBI Data “Sliced,” or binary, VBI data is commonly available from NTSC/PAL video decoders. This has the advantage of lower data rates since binary, rather than oversampled, data is present. The primary disadvantage is the variety of techniques NTSC/PAL video decoder chip manufacturers use to transfer the “sliced” VBI data over the video interface.
NTSC/PAL Decoder Considerations Closed Captioning In addition to caption and text commands that clear the display, five other events typically force the display to be cleared: (1) A change in the caption display mode, such as switching from CC1 to T1. (2) A loss of video lock, such as on a channel change, forces the display to be cleared. The currently active display mode does not change. For example, if CC1 was selected before loss of video lock, it remains selected.
389
(3) Activation of autoblanking. If the caption signal has not been detected for (typically) 32 consecutive frames, or no new data for the selected channel has been received for (typically) 512 frames, the display memory is cleared. Once the caption signal has been detected for (typically) 32 consecutive frames, or new data has been received, it is displayed. (4) A “clear” command (from the remote control for example) forces the display to be cleared. (5) Disabling caption decoding also forces the display to be cleared.
Widescreen Signaling The decoder must be able to handle a variety of WSS inputs including: (1) PAL or NTSC WSS signal on composite, svideo or Y (of YPbPr). (2) SCART analog inputs (DC offset indicator) (3) S-Video analog inputs (DC offset indicator)
In addition to automatically processing the video signal to fit a 4:3 or 16:9 display based on the WSS data, the decoder should also support manual overrides in case the user wishes a specific mode of operation due to personal preferences. Software uses this aspect ratio information, user preferences, and display format to assist in properly processing the program for display.
390
Chapter 8: NTSC, PAL, and SECAM Overview
Ghost Cancellation Ghost cancellation (the removal of undesired reflections present in the signal) is required due to the high data rate of some services, such as teletext. Ghosting greater than 100 ns and –12 dB corrupts teletext data. Ghosting greater than –3 dB is difficult to remove cost-effectively in hardware or software, while ghosting less than –12 dB need not be removed. Ghost cancellation for VBI data is not as complex as ghost cancellation for active video. Unfortunately, the GCR (ghost cancellation reference) signal is not usually present. Thus, a ghost cancellation algorithm must determine the amount of ghosting using other available signals, such as the serration and equalization pulses. The NTSC GCR signal is specified in ATSC A/49 and ITU-R BT.1124. If present, it occupies lines 19 and 282. The GCR permits the detection of ghosting from –3 to +45 µs, and follows an 8-field sequence. The PAL GCR signal is specified in BT.1124 and ETSI ETS 300 732. If present, it occupies line 318. The GCR permits the detection of ghosting from –3 to +45 µs, and follows a 4-frame sequence.
Enhanced Television Programming The enhanced television programming standard (SMPTE 363M) is used for creating and delivering enhanced and interactive programs. The enhanced content can be delivered over a variety of mediums—including analog and digital television broadcasts—using terrestrial, cable and satellite networks. In defining how to create enhanced content, the specifica-
tion defines the minimum receiver functionality. To minimize the creation of new specifications, it leverages Internet technologies such as HTML and Java-script. The benefits of doing this are that there are already millions of pages of potential content, and the ability to use existing web-authoring tools. The specification mandates that receivers support, as a minimum, HTML 4.0, Javascript 1.1, and Cascading Style Sheets. Supporting additional capabilities, such as Java and VRML, is optional. This ensures content is available to the maximum number of viewers. For increased capability, a new “tv:” attribute is added to the HTML. This attribute enables the insertion of the television program into the content, and may be used in a HTML document anywhere that a regular image may be placed. Creating an enhanced content page that displays the current television channel anywhere on the display is as easy as inserting an image in a HTML document. The specification also defines how the receivers obtain the content and how they are informed that enhancements are available. The latter task is accomplished with triggers.
Triggers Triggers alert receivers to content enhancements, and contain information about the enhancements. Among other things, triggers contain a Universal Resource Locator (URL) that defines the location of the enhanced content. Content may reside locally—such as when delivered over the network and cached to a local hard drive—or it may reside on the Internet or another network. Triggers may also contain a human-readable description of the content. For example, it may contain the description “Press ORDER to order this product,” which can be displayed for
Enhanced Television Programming
the viewer. Triggers also may contain expiration information, indicating how long the enhancement should be offered to the viewer. Lastly, triggers may contain scripts that trigger the execution of Javascript within the associated HTML page, to support synchronization of the enhanced content with the video signal and updating of dynamic screen data. The processing of triggers is defined in SMPTE 363M and is independent of the method used to carry them.
Transports Besides defining how content is displayed and how the receiver is notified of new content, the specification also defines how content is delivered. Because a receiver may not have an Internet connection, the specification describes two models for delivering content. These two models are called transports, and the two transports are referred to as Transport Type A and Transport Type B. If the receiver has a back-channel (or return path) to the Internet, Transport Type A will broadcast the trigger and the content will be pulled over the Internet. If the receiver does not have an Internet connection, Transport Type B provides for delivery of both triggers and content via the broadcast medium. Announcements are sent over the network to associate triggers with content streams. An announcement describes the content, and may include information regarding bandwidth, storage requirements, and language.
Delivery Protocols For traditional bi-directional Internet communication, the Hypertext Transfer Protocol (HTTP) defines how data is transferred at the application level. For uni-directional broad-
391
casts where a two-way connection is not available, SMPTE 364M defines a uni-directional application-level protocol for data delivery: Uni-directional Hypertext Transfer Protocol (UHTTP). Like HTTP, UHTTP uses traditional URL naming schemes to reference content. Content can reference enhancement pages using the standard “http:” and “ftp:” naming schemes. A “lid:,” or local identifier, URL is also available to allow reference to content that exists locally (such as on the receiver’s hard drive) as opposed to on the Internet or other network.
Bindings How data is delivered over a specific network is called “binding.” Bindings have been defined for NTSC and PAL. NTSC Bindings Transport Type A triggers are broadcast on data channel 2 of the EIA-608 captioning signal. Transport Type B binding also includes a mechanism for delivering IP multicast packets over the vertical blanking interval (VBI), otherwise known as IP over VBI (IP/VBI). At the lowest level, the television signal transports NABTS (North American Basic Teletext Standard) packets during the VBI. These NABTS packets are recovered to form a sequential data stream (encapsulated in a SLIP-like protocol) that is unframed to produce IP packets. PAL Bindings Both transport types are based on carriage of IP multicast packets in VBI lines of a PAL system by means of teletext packets 30 or 31. Transport Type A triggers are carried in UDP/IP multicast packets, delivered to address 224.0.23.13 and port 2670.
392
Chapter 8: NTSC, PAL, and SECAM Overview
Transport Type B (described in SMPTE 357M) carries a single trigger in a single UDP/IP multicast packet, delivered on the address and port defined in the SDP announcement for the enhanced television program. The trigger protocol is very lightweight in order to provide quick synchronization.
References 1. Advanced Television Enhancement Forum, Enhanced Content Specification, 1999. 2. ATSC A/49, 13 May 1993, Ghost Cancelling Reference Signal for NTSC. 3. BBC Technical Requirements for Digital Television Services, Version 1.0, February 3, 1999, BBC Broadcast. 4. EIA–189–A, July 1976, Encoded Color Bar Signal. 5. EIA–516, May 1988, North American Basic Teletext Specification (NABTS). 6. EIA–608, September 1994, Recommended Practice for Line 21 Data Service. 7. EIA-J CPR–1204, Transfer Method of Video ID Information using Vertical Blanking Interval (525-line System), March 1997. 8. ETSI EN 300 163, Television Systems: NICAM 728: Transmission of Two Channel Digital Sound with Terrestrial Television Systems B, G, H, I, K1, and L, March 1998. 9. ETSI EN 300 231, Television Systems: Specification of the Domestic Video Programme Delivery Control System (PDC), December 2002. 10. ETSI EN 300 294, Television Systems: 625line Television Widescreen Signalling (WSS), December 2002. 11. ETSI EN 300 706, Enhanced Teletext Specification, December 2002.
12. ETSI EN 300 708, Television Systems: Data Transmission within Teletext, December 2002. 13. ETSI ETS 300 731, Television Systems: Enhanced 625-Line Phased Alternate Line (PAL) Television: PALplus, March 1997. 14. ETSI ETS 300 732, Television Systems: Enhanced 625-Line PAL/SECAM Television; Ghost Cancellation Reference (GCR) Signals, January 1997. 15. Faroudja, Yves Charles, NTSC and Beyond, IEEE Transactions on Consumer Electronics, Vol. 34, No. 1, February 1988. 16. IEC 61880, 1998–1, Video Systems (525/ 60)—Video and Accompanied Data Using the Vertical Blanking Interval—Analog Interface. 17. ITU-R BS.707–3, 1998, Transmission of Multisound in Terrestrial Television Systems PAL B, G, H, and I and SECAM D, K, K1, and L. 18. ITU-R BT.470–6, 1998, Conventional Television Systems. 19. ITU-R BT.471–1, 1986, Nomenclature and Description of Colour Bar Signals. 20. ITU-R BT.472–3, 1990, Video Frequency Characteristics of a Television System to Be Used for the International Exchange of Programmes Between Countries that Have Adopted 625-Line Colour or Monochrome Systems. 21. ITU-R BT.473–5, 1990, Insertion of Test Signals in the Field-Blanking Interval of Monochrome and Colour Television Signals. 22. ITU-R BT.569–2, 1986, Definition of Parameters for Simplified Automatic Measurement of Television Insertion Test Signals. 23. ITU-R BT.653–3, 1998, Teletext Systems. 24. ITU-R BT.809, 1992, Programme Delivery Control (PDC) System for Video Recording.
References
25. ITU-R BT.1118, 1994, Enhanced Compatible Widescreen Television Based on Conventional Television Systems. 26. ITU-R BT.1119–2, 1998, Wide-Screen Signalling for Broadcasting. 27. ITU-R BT.1124, 1994, Reference Signals for Ghost Cancelling in Analogue Television Systems. 28. ITU-R BT.1197–1, 1998, Enhanced WideScreen PAL TV Transmission System (the PALplus System). 29. ITU-R BT.1298, 1997, Enhanced WideScreen NTSC TV Transmission System. 30. Multichannel TV Sound System BTSC System Recommended Practices, EIA Television Systems Bulletin No. 5 (TVSB5). 31. NTSC Video Measurements, Tektronix, Inc., 1997. 32. SMPTE 12M–1999, Television, Audio and Film—Time and Control Code. 33. SMPTE 170M–1999, Television—Composite Analog Video Signal—NTSC for Studio Applications. 34. SMPTE 262M–1995, Television, Audio and Film—Binary Groups of Time and Control Codes—Storage and Transmission of Data. 35. SMPTE 309M–1999, Television—Transmission of Date and Time Zone Information in Binary Groups of Time and Control Code.
393
36. SMPTE 357M–2002, Television—Declarative Data Essence -- Internet Protocol Multicast Encapsulation. 37. SMPTE 361M–2002, Television—NTSC IP and Trigger Binding to VBI. 38. SMPTE 363M–2002, Television—Declarative Data Essence -- Content Level 1. 39. SMPTE 364M–2001, Declarative Data Essence -- Unidirectional Hypertext Transport Protocol. 40. SMPTE RP-164–1996, Location of Vertical Interval Time Code. 41. SMPTE RP-186–1995, Video Index Information Coding for 525- and 625-Line Television Systems. 42. SMPTE RP-201–1999, Encoding Film Transfer Information Using Vertical Interval Time Code. 43. Specification of Television Standards for 625-Line System-I Transmissions, 1971, Independent Television Authority (ITA) and British Broadcasting Corporation (BBC). 44. Television Measurements, NTSC Systems, Tektronix, Inc., 1998. 45. Television Measurements, PAL Systems, Tektronix, Inc., 1990.
394
Chapter 9: NTSC/PAL Digital Encoding and Decoding
Chapter 9: NTSC and PAL Digital Encoding and Decoding Chapter 9
NTSC and PAL Digital Encoding and Decoding Although not exactly “digital” video, the NTSC and PAL composite color video formats are currently the most common formats for video. Although the video signals themselves are analog, they can be encoded and decoded almost entirely digitally. Analog NTSC and PAL encoders and decoders have been available for some time. However, they have been difficult to use, required adjustment, and offered limited video quality. Using digital techniques to implement NTSC and PAL encoding and decoding offers many advantages such as ease of use, minimum analog adjustments, and excellent video quality.
394
In addition to composite video, S-video is supported by consumer and pro-video equipment, and should also be implemented. S-video uses separate luminance (Y) and chrominance (C) analog video signals so higher quality may be maintained by eliminating the Y/C separation process. This chapter discusses the design of a digital encoder (Figure 9.1) and decoder (Figure 9.21) that support composite and S-video (M) NTSC and (B, D, G, H, I, NC) PAL video signals. (M) and (N) PAL are easily accommodated with some slight modifications. NTSC encoders and decoders are usually based on the YCbCr, YUV, or YIQ color space. PAL encoders and decoders are usually based on the YCbCr or YUV color space.
NTSC and PAL Encoding
Video Standard
Sample Clock Rate
Applications
9 MHz
SVCD
13.5 MHz
720 × 480i
12.27 MHz
square pixels
9 MHz
SVCD
14.75 MHz
square pixels
13.5 MHz
MPEG-2 DV
572 × 525i
704 × 480i
MPEG-2
BT.601
480 × 480i
720 × 480i
DV
(B, D, G, H, I, N, NC) PAL
Total Resolution
1
BT.601 (M) NTSC, (M) PAL
Active Resolution
640 × 480i
480 × 576i 768 × 576i
720 × 576i 2
704 × 576i
720 × 576i
858 × 525i
395
Field Rate (per second)
59.94 interlaced
780 × 525i
576 × 625i
944 × 625i 864 × 625i
50 interlaced
Table 9.1. Common NTSC/PAL Sample Rates and Resolutions. 1Typically 716 true active samples between 10% blanking points. 2Typically 702 true active samples between 50% blanking points.
NTSC and PAL Encoding YCbCr input data has a nominal range of 16–235 for Y and 16–240 for Cb and Cr. RGB input data has a range of 0–255; pro-video applications may use a nominal range of 16–235. As YCbCr values outside these ranges result in overflowing the standard YIQ or YUV ranges for some color combinations, one of three things may be done, in order of preference: (a) allow the video signal to be generated using the extended YIQ or YUV ranges; (b) limit the color saturation to ensure a legal video signal is generated; or (c) clip the YIQ or YUV levels to the valid ranges. 4:1:1, 4:2:0, or 4:2:2 YCbCr data must be converted to 4:4:4 YCbCr data before being converted to YIQ or YUV data. The chrominance lowpass filters will not perform the interpolation properly. Table 9.1 lists some of the common sample rates and resolutions.
2× Oversampling 2× oversampling generates 8:8:8 YCbCr or RGB data, simplifying the analog output filters. The oversampler is also a convenient place to convert from 8-bit to 10-bit data, providing an increase in video quality.
Color Space Conversion Choosing the 10-bit video levels to be white = 800 and sync = 16, and knowing that the sync-to-white amplitude is 1V, the full-scale output of the D/A converters (DACs) is therefore set to 1.305V. (M) NTSC, (M, N) PAL Since (M) NTSC and (M, N) PAL have a 7.5 IRE blanking pedestal and a 40 IRE sync amplitude, the color space conversion equations are derived so as to generate 0.660V of active video.
396
Chapter 9: NTSC/PAL Digital Encoding and Decoding
HSYNC# VSYNC#
VIDEO TIMING AND GENLOCK CONTROL
BLANK# FIELD_0 FIELD_1 CLOCK
BLANK PEDESTAL
Y
CR
CB
2X OVERSAMPLE
BLANK RISE / FALL EXPANDER
SYNC RISE / FALL EXPANDER
+
+
+
2X OVERSAMPLE ---------1.3 MHZ LPF
Y
DAC
NTSC / PAL
DAC
C
+
MUX
2X OVERSAMPLE ---------1.3 MHZ LPF
DAC
MUX
SIN ROM BURST CONTROL
COS ROM
DTO
Figure 9.1. Typical NTSC/PAL Digital Encoder Implementation.
NTSC and PAL Encoding
YUV Color Space Processing Modern encoder designs are now based on the YUV color space, For these encoders, the YCbCr to YUV equations are: Y = 0.591(Y601 – 64)
Y = 0.591(Y601 – 64) I = 0.596(Cr – 512) – 0.274(Cb – 512) Q = 0.387(Cr – 512) + 0.423(Cb – 512)
The R´G´B´ to YIQ equations are:
U = 0.504(Cb – 512)
Y = 0.151R´ + 0.297G´ + 0.058B´
V = 0.711(Cr – 512)
I = 0.302R´ – 0.139G´ – 0.163B´
The R´G´B´ to YUV equations are: Y = 0.151R´ + 0.297G´ + 0.058B´ U = –0.074R´ – 0.147G´ + 0.221B´ V = 0.312R´ – 0.261G´ – 0.051B´
For pro-video applications using a 10-bit nominal range of 64–940 for RGB, the R´G´B´ to YUV equations are: Y = 0.177(R´ – 64) + 0.347(G´ – 64) + 0.067(B´ – 64) U = –0.087(R´ – 64) – 0.171(G´ – 64) + 0.258(B´ – 64) V = 0.364(R´ – 64) – 0.305(G´ – 64) – 0.059(B´ – 64)
Y has a nominal range of 0 to 518, U a nominal range of 0 to ±226, and V a nominal range of 0 to ±319. Negative values of Y should be supported to allow test signals, keying information, and real-world video to be passed through the encoder with minimum corruption. YIQ Color Space Processing For older NTSC encoder designs based on the YIQ color space, the YCbCr to YIQ equations are:
397
Q = 0.107R´ – 0.265G´ + 0.158B´
For pro-video applications using a 10-bit nominal range of 64–940 for R´G´B´, the R´G´B´ to YIQ equations are: Y = 0.177(R´ – 64) + 0.347(G´ – 64) + 0.067(B´ – 64) I = 0.352(R´ – 64) – 0.162(G´ – 64) – 0.190(B´ – 64) Q = 0.125(R´ – 64) – 0.309(G´ – 64) + 0.184(B´ – 64)
Y has a nominal range of 0 to 518, I a nominal range of 0 to ±309, and Q a nominal range of 0 to ±271. Negative values of Y should be supported to allow test signals, keying information, and real-world video to be passed through the encoder with minimum corruption. YCbCr Color Space Processing If the design is based on the YUV color space, the Cb and Cr conversion to U and V may be avoided by scaling the sin and cos values during the modulation process or scaling the color difference lowpass filter coefficients. This has the advantage of reducing data path processing.
398
Chapter 9: NTSC/PAL Digital Encoding and Decoding
NTSC–J Since the version of (M) NTSC used in Japan has a 0 IRE blanking pedestal, the color space conversion equations are derived so as to generate 0.714V of active video. YUV Color Space Processing The YCbCr to YUV equations are: Y = 0.639(Y601 – 64) U = 0.545(Cb – 512) V = 0.769(Cr – 512)
The R´G´B´ to YUV equations are: Y = 0.164R´ + 0.321G´ + 0.062B´ U = –0.080R´ – 0.159G´ + 0.239B´ V = 0.337R´ – 0.282G´ – 0.055B´
For pro-video applications using a 10-bit nominal range of 64–940 for R´G´B´, the R´G´B´ to YUV equations are: Y = 0.191(R´ – 64) + 0.375(G´ – 64) + 0.073(B´ – 64) U = –0.094(R´ – 64) – 0.185(G´ – 64) + 0.279(B´ – 64) V = 0.393(R´– 64) – 0.329(G´ – 64) – 0.064(B´ – 64)
Y has a nominal range of 0 to 560, U a nominal range of 0 to ±244, and V a nominal range of 0 to ±344. Negative values of Y should be supported to allow test signals, keying information, and real-world video to be passed through the encoder with minimum corruption. YIQ Color Space Processing For older encoder designs based on the YIQ color space, the YCbCr to YIQ equations are:
Y = 0.639(Y601 – 64) I = 0.645(Cr – 512) – 0.297(Cb – 512) Q = 0.419(Cr – 512) + 0.457(Cb – 512)
The R´G´B´ to YIQ equations are: Y = 0.164R´ + 0.321G´ + 0.062B´ I = 0.326R´ – 0.150G´ – 0.176B´ Q = 0.116R´ – 0.286G´ + 0.170B´
For pro-video applications using a 10-bit nominal range of 64–940 for R´G´B´, the R´G´B´ to YIQ equations are: Y = 0.191(R´ – 64) + 0.375(G´ – 64) + 0.073(B´ – 64) I = 0.381(R´ – 64) – 0.176(G´ – 64) – 0.205(B´ – 64) Q = 0.135(R´ – 64) – 0.334(G´ – 64) + 0.199(B´ – 64)
Y has a nominal range of 0 to 560, I a nominal range of 0 to ±334, and Q a nominal range of 0 to ±293. Negative values of Y should be supported to allow test signals, keying information, and real-world video to be passed through the encoder with minimum corruption. YCbCr Color Space Processing If the design is based on the YUV color space, the Cb and Cr conversion to U and V may be avoided by scaling the sin and cos values during the modulation process or scaling the color difference lowpass filter coefficients. This has the advantage of reducing data path processing.
NTSC and PAL Encoding
(B, D, G, H, I, NC) PAL Since these PAL standards have a 0 IRE blanking pedestal and a 43 IRE sync amplitude, the color space conversion equations are derived so as to generate 0.7V of active video. YUV Color Space Processing The YCbCr to YUV equations are: Y = 0.625(Y601 – 64) U = 0.533(Cb – 512) V = 0.752(Cr – 512)
The R´G´B´ to YUV equations are: Y = 0.160R´ + 0.314G´ + 0.061B´ U = –0.079R´ – 0.155G´ + 0.234B´ V = 0.329R´ – 0.275G´ – 0.054B´
For pro-video applications using a 10-bit nominal range of 64–940 for R´G´B´, the R´G´B´ to YUV equations are: Y = 0.187(R´ – 64) + 0.367(G´ – 64) + 0.071(B´ – 64) U = –0.092(R´ – 64) – 0.181(G´ – 64) + 0.273(B´ – 64) V = 0.385(R´– 64) – 0.322(G´ – 64) – 0.063(B´ – 64)
Y has a nominal range of 0 to 548, U a nominal range of 0 to ±239, and V a nominal range of 0 to ±337. Negative values of Y should be supported to allow test signals, keying information, and real-world video to be passed through the encoder with minimum corruption. YCbCr Color Space Processing If the design is based on the YUV color space, the Cb and Cr conversion to U and V may be avoided by scaling the sin and cos val-
399
ues during the modulation process or scaling the color difference lowpass filter coefficients. This has the advantage of reducing data path processing.
Luminance (Y) Processing Lowpass filtering to about 6 MHz must be done to remove high-frequency components generated as a result of the 2x oversampling process. An optional notch filter may also be used to remove the color subcarrier frequency from the luminance information. This improves decoded video quality for decoders that use simple Y/C separation. The notch filter should be disabled when generating S-video, RGB, or YPbPr video signals. Next, any blanking pedestal is added during active video, and the blanking and sync information are added. (M) NTSC, (M, N) PAL As (M) NTSC and (M, N) PAL have a 7.5 IRE blanking pedestal, a value of 42 is added to the luminance data during active video. 0 is added during the blank time. After the blanking pedestal is added, the luminance data is clamped by a blanking signal that has a raised cosine distribution to slow the slew rate of the start and end of the video signal. Typical blank rise and fall times are 140 ±20 ns for NTSC and 300 ±100 ns for PAL. Digital composite sync information is added to the luminance data after the blank processing has been performed. Values of 16 (sync present) or 240 (no sync) are assigned. The sync rise and fall times should be processed to generate a raised cosine distribution (between 16 and 240) to slow the slew rate of the sync signal. Typical sync rise and fall times are 140 ±20 ns for NTSC and 250 ±50 ns for
400
Chapter 9: NTSC/PAL Digital Encoding and Decoding
PAL, although the encoder should generate sync edges of about 130 or 240 ns to compensate for the analog output filters slowing the sync edges. At this point, we have digital luminance with sync and blanking information, as shown in Table 9.2.
compensate for the analog output filters slowing the sync edges. At this point, we have digital luminance with sync and blanking information, as shown in Table 9.2. Analog Luminance (Y) Generation The digital luminance data may drive a 10bit DAC that generates a 0–1.305V output to generate the Y video signal of a S-video (Y/C) interface. Figures 9.2 and 9.3 show the luminance video waveforms for 75% color bars. The numbers on the luminance levels indicate the data value for a 10-bit DAC with a full-scale output value of 1.305V. The video signal at the connector should have a source impedance of 75Ω. As the sample-and-hold action of the DAC introduces a (sin x)/x characteristic, the video data may be digitally filtered by a [(sin x)/x]–1 filter to compensate. Alternately, as an analog lowpass filter is usually present after the DAC, the correction may take place in the analog filter. As an option, the ability to delay the digital Y information a programmable number of clock cycles before driving the DAC may be useful. If the analog luminance video is lowpass filtered after the DAC, and the analog chrominance video is bandpass filtered after its
NTSC–J When generating NTSC–J video, there is a 0 IRE blanking pedestal. Thus, no blanking pedestal is added to the luminance data during active video. Otherwise, the processing is the same as for (M) NTSC. (B, D, G, H, I, NC) PAL When generating (B, D, G, H, I, NC) PAL video, there is a 0 IRE blanking pedestal. Thus, no blanking pedestal is added to the luminance data during active video. Blanking information is inserted using the same technique as used for (M) NTSC. However, typical blank rise and fall times are 300 ±100 ns. Composite sync information is added using the same technique as used for (M) NTSC, except values of 16 (sync present) or 252 (no sync) are used. Typical sync rise and fall times are 250 ±50 ns, although the encoder should generate sync edges of about 240 ns to
Video Level
(M) NTSC
NTSC–J
(B, D, G, H, I, NC) PAL
(M, N) PAL
white
800
800
800
800
black
282
240
252
282
blank
240
240
252
240
sync
16
16
16
16
Table 9.2. 10-Bit Digital Luminance Values.
1.020 V
401
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
WHITE
NTSC and PAL Encoding
WHITE LEVEL (800)
671 626 554 100 IRE
510 442 398 326
0.357 V
BLACK LEVEL (282)
7.5 IRE
0.306 V
BLANK LEVEL (240)
40 IRE
0.020 V
SYNC LEVEL (16)
800
1.020 V
BLACK
BLUE
RED
MAGENTA
GREEN
CYAN
YELLOW
WHITE
Figure 9.2. (M) NTSC Luminance (Y) Video Signal for 75% Color Bars. Indicated luminance levels are 10-bit values.
WHITE LEVEL (800)
616
100 IRE
540 493 422 375
299 0.321 V
BLACK / BLANK LEVEL (252)
43 IRE
0.020 V
SYNC LEVEL (16)
Figure 9.3. (B, D, G, H, I) PAL Luminance (Y) Video Signal for 75% Color Bars. Indicated luminance levels are 10-bit values.
402
Chapter 9: NTSC/PAL Digital Encoding and Decoding
DAC, the chrominance video path may have a longer delay (typically up to about 400 ns) than the luminance video path. By adjusting the delay of the Y data, the analog luminance and chrominance video will be aligned more closely after filtering, simplifying the analog design.
Color Difference Processing Lowpass Filtering The color difference signals (CbCr, UV, or IQ) should be lowpass filtered using a Gaussian filter. This filter type minimizes ringing and overshoot, avoiding the generation of visual artifacts on sharp edges. If the encoder is used in a video editing application, the filters should have a maximum ripple of ±0.1 dB in the passband. This minimizes the cumulation of gain and loss artifacts due to the filters, especially when multiple passes through the encoding and decoding processes are done. At the final encoding point, Gaussian filters may be used. YCbCr and YUV Color Space Cb and Cr, or U and V, are lowpass filtered to about 1.3 MHz. Typical filter characteristics are 20 dB attenuation at 3.6 MHz. The filter characteristics are shown in Figure 9.4. YIQ Color Space Q is lowpass filtered to about 0.6 MHz. Typical filter characteristics are